Thursday, 2015-06-18

jheskethjeblair: I believe it's the duration of git pack upload00:00
jheskethspecifically for projects that have gotten very large00:00
cathy__jeblair: thanks for the info. So how do we announce the nomination period? In the IRC meeting? how long should the nomination period be?00:00
jeblaircathy__: usually a week, and usually over email00:01
cathy__jeblair: over the openstack-dev or among the contributors?00:02
*** sdake has quit IRC00:02
cathy__anteaya: jhesketh : thanks for the link. I will go through it00:03
jeblaircathy__: openstack-dev00:03
cathy__jeblair: ok, thanks!00:03
*** esmute has joined #openstack-infra00:04
* SpamapS is back00:04
SpamapSgreghaynes: weird, vsp still says used00:04
greghaynesSpamapS: for undercloud?00:05
greghaynesSpamapS: or for your node?00:06
jeblairjhesketh: those look like idle timeouts, so if they have an effect, it will be because either the client or server are unresponsive for > 60 seconds00:07
jeblairjhesketh: (i think that's also the case for apache)00:07
SpamapSgreghaynes: for my node00:08
jheskethjeblair: right, okay.. so wouldn't upping those haproxy timeouts be helpful then?00:08
* jhesketh is not great with networking00:08
openstackgerritJohn L. Villalovos proposed openstack-infra/irc-meetings: Add href URL for the IRC channel in template
jeblairjhesketh: yes, if it's taking the server longer than 60 seconds to even start responding to the request.00:09
jheskethjeblair: would the queue time have anything to do with it?00:10
jeblairjhesketh: (and if it is, we need to improve the git server farm)00:10
jeblairjhesketh: i would not expect it to00:10
jheskethapparently quite a few people have been hitting this00:10
*** cathy__ has quit IRC00:10
jheskethwe've been hitting it for a long time, but it's become increasingly common in the last few weeks00:10
jeblairthere's a huge amount of really good logging on the haproxy side00:10
jhesketh(it also took me a long time to figure out what was going on)00:10
SpamapSgreghaynes: I was able to get textcons working00:10
jeblairif you have a timestamp + ip, we can look it up00:11
jheskethjeblair: those logs aren't published anywhere are they00:11
jeblairjhesketh: nope00:11
*** tlian has joined #openstack-infra00:11
jeblairjhesketh: but if you have ip+timestamp i'll be happy to look em up00:11
jheskethjeblair: aren't the timestamps in that dump?00:11
jheskethand IP depending on which node you're looking at00:12
jeblairjhesketh: i was hoping you weren't going to make me sort through packet traces :)00:12
jheskethjeblair: haha00:12
jheskethjeblair: okay, let me pick one and give you some data00:12
greghaynesSpamapS: ipmitool is working fine for me now btw, so I can assert state with it on things00:12
SpamapSgreghaynes: good00:13
SpamapSoh interesting00:13
SpamapSthe MAC you had for that box is listed as no link00:13
SpamapSwhich is consistent with what I remember.. I think we want net100:14
greghaynesSpamapS: ive noticed it to be not totally consistent00:14
SpamapSgreghaynes: that sucks00:14
SpamapSit was consistent in hp1 for a while at least ;)00:14
greghayneswhich is one of the big things that motivated me to get an inventory with all the macs for each host00:14
SpamapSwait no, net0 did have it00:15
SpamapSmy dnsmasq just isn't seeing it00:15
SpamapSgah, and these are setup "if you don't hear from DHCP, just sit there like a dope"00:15
SpamapSinstead of retrying dhcp forever00:15
jheskethjeblair: ip is, time began around Date: Wed, 17 Jun 2015 10:02:04 GMT00:15
SpamapSgreghaynes: net1 has link, and is the MAC you gave me. I was just going cross-eyed. ;)00:16
jeblairjhesketh: thanks.  good timing; i just dug up the haproxy log reference00:16
greghaynesSpamapS: ok, im ready to ipmi power off everything but your box00:17
greghaynespunch it?00:17
jeblairJun 17 10:03:01 localhost haproxy[1404]: [17/Jun/2015:10:02:04.786] balance_git_http balance_git_http/ 1/0/56428 6239845 -- 70/15/15/4/0 0/000:17
jeblairJun 17 10:04:43 localhost haproxy[1404]: [17/Jun/2015:10:03:02.697] balance_git_http balance_git_http/ 1/0/100997 311 cD 63/14/14/7/0 0/000:17
SpamapSgreghaynes: double check you're not powering off the bastion, but then yeha, punch it00:17
greghaynesgood call00:17
jeblairjhesketh: 2 entries for that; the second says "the client-side timeout expired while waiting for the client to send or receive data." and "the session was in the DATA phase."00:18
jeblairjhesketh: (that's what the cD mean)00:18
jeblairjhesketh: the first one was normal00:18
lifelessjeblair: is now green00:18
lifelessjeblair: and has the pbr ordering fix in it00:18
greghaynesSpamapS: you can double check im right - bastion is ipmi IP
jeblairlifeless: woot, thx00:19
lifelessmtreinish: ^ can you review too please?00:19
SpamapSgreghaynes: looking00:19
greghaynesSpamapS: actually, you can just /home/shared/ - its set up to echo out a dry run00:20
greghaynesSpamapS: for a bit of code review00:20
jheskethjeblair: the other thing I was going to do after fixing the timeouts was check we're repacking the git repositories appropriately (which we may already be doing)00:20
SpamapSgreghaynes: yeah you're good00:20
jeblairjhesketh: the way i read this is that there is a problem between haproxy and the client.00:20
*** ashleighfarnham has quit IRC00:20
openstackgerritBradley Jones proposed openstack-infra/project-config: Create Magnum UI repo
jheskethjeblair: agreed, and the patch I pushed should fix that no?00:21
jheskeththe git pack upload is just too large at the moment00:21
jeblairjhesketh: too large for what?  these are idle timeouts, meaning that no data are being transferred.00:21
jheskeththe request is taking too long to send (if I'm understanding it correctly)00:22
*** yolanda has quit IRC00:22
jheskeththe git client takes too long asking the git server for all the packs it wants00:23
*** gyee has quit IRC00:23
jeblairjhesketh: as long as data are moving, i don't think these timeout should be hit00:23
*** yamamoto has quit IRC00:23
*** yolanda has joined #openstack-infra00:23
jeblairjhesketh: i think increasing the timeout will reduce or mask the appearance of the problem, but we won't be closer to understanding what's really going on.00:23
*** hichihara has joined #openstack-infra00:24
jeblairjhesketh: if you want to do that, i'm okay with that, but i also think that running cordinated tcpdumps on a remote system as well as an haproxy node would tell us something.00:24
jeblairjhesketh: btw, which OS and protocol were you testing?00:25
lifelessquick question, and sorry if this is in the docs. If A Depends-On B (different repos), and B is updated, is A retested automatically ?00:25
*** ashleighfarnham has joined #openstack-infra00:25
*** puranamr has joined #openstack-infra00:25
jheskethjeblair: hmm okay, I perhaps don't understand enough of networking here to comment further.. This was gus's take on it earlier though if it helps
jeblairlifeless: i don't know off the top of my head00:26
jheskethjeblair: ubuntu 12.04, http. Although very similar failures on https and git://00:26
*** zz_dimtruck is now known as dimtruck00:28
SpamapShrm, dnsmasq is failing me. may have to switch to isc-dhcpd just because I'm o_O on why it isn't responding. :-P00:29
greghaynesSpamapS: I tend to isc by default, I find it to do the things I want it to more often00:29
*** rguillebert has quit IRC00:29
SpamapSgreghaynes: yeah I just figured.. dnsmasq is there.. its supposed to do this.. :-P00:30
*** ashleighfarnham has quit IRC00:30
*** rossella_s has quit IRC00:30
greghaynesSpamapS: ok, so there are two hosts that do not care that im telling them chassis power off, they still report power on00:30
greghaynesother than that they are all off00:30
greghaynesim guessing those two have the ilo lockup issue or similar00:30
SpamapSgreghaynes: try ssh'ing into the ilo, if that fails, hit the web ilo interface and reset the ilo00:31
*** rossella_s has joined #openstack-infra00:31
SpamapSgreghaynes: note the machines, and the ilo version, and lets see if the noc can upgrade them.00:31
greghaynesyea, server power: reset00:31
greghaynesour old friend00:31
*** Longgeek has joined #openstack-infra00:32
SpamapSgreghaynes: I'm going to shut down the seed too00:33
SpamapSI think it is interfering00:33
*** tlian has quit IRC00:34
*** jamielennox|away is now known as jamielennox00:35
lifelessok -> getting some fresh air; SMS / hangouts if I'm needed.00:35
*** clu_ has quit IRC00:35
SpamapSdnsmasq not binding to dhcp port00:35
jeblairjhesketh: looking at the graphs, i believe you did DoS the git server you were gitting (04)00:36
jeblairjhesketh: so it's reasonable to suspect that it did not respond to a request in <1m.  i'm not sure why haproxy would account for that as a client timeout, but it's the same as a server timeout, so maybe that's just what it hit first. (the docs recommend keeping them the same)00:37
jheskethjeblair: I can't see the DoS, what are you looking at?00:38
jeblairjhesketh: at any rate, i do think your change will mask the error, but i also think that we should consider tuning haproxy, perhaps reducing the number of concurrent connections it send to backends.  unfortunately, since we round-robin between two, their accounting is wrong.  so we may actually need to split up the pool and give half to one server and half to another.00:38
greghaynesSpamapS: ok, reset had no effect - noted them in the machine-information-tab.txt00:39
*** jasondotstar has quit IRC00:39
jheskethand if we are hitting the limits of our git pool, we should look at updating them because turbo-hipster is regularly failing to clone and I could reproduce the problem cloning across only 15 hosts (possibly even less)00:39
jheskethactually I know people who have had the problem just doing a single clone at the wrong time00:39
greghaynesSpamapS: One of them was already in there as "ILO hung, no ticket filed yet"00:39
clarkbjhesketh jeblair I thibk nova has outgrown the bacjends00:39
SpamapSgreghaynes: nice00:39
jeblairjhesketh: (this is all actually a little maddening; we don't need two hosts except that we're hitting rax's bandwidth limits for vms, and we can't use rax's own load balancer service because it doesn't expose the right options; perhaps we should see if onmetal would fix us)00:40
clarkbnova operations consume almost a gig of memory last I saw and those pile up00:40
clarkband yes lbaas is very http centric so not great for git://00:40
jeblairclarkb: i think that's why i suggested that we have haproxy queue and reduce concurrent backend connections?00:40
clarkbjeblair ah ok, we could also make bacjends bigger00:40
clarkbor make nova smaller00:41
jeblairi'd rather scale out than up00:41
*** tlian has joined #openstack-infra00:41
jeblairthe machines are mostly underpowered, until a bunch of simultaneous nova clones happen00:41
jeblairjhesketh: look at the load average, memory usage, and swap graphs00:42
SpamapSah, I told dnsmasq to only do ipv6. silly00:43
jeblairi think that establishing the number of simultaneous connections a backend can support and then configuring haproxy for that will keep the backends from being overloaded, then we just add more backends to keep the haproxy queue shallow00:43
*** dims has joined #openstack-infra00:43
jheskethjeblair: ah, I didn't think they lined up with the tcpdumps but that was me timezone failing (wireshark converted the tz)00:44
jeblairjhesketh, clarkb: how about an onmetal compute node for haproxy?00:44
jheskethcan't hurt00:44
*** otter768 has quit IRC00:44
clarkbjeblair what is the bw limit for those?00:45
jeblairthat will get us back to a single one so the accounting is correct00:45
jeblair"Redundant 10Gbps network connections"; i don't see anything about a limit, but maybe i haven't found it yet?00:45
jeblairjhesketh: i'm cool with the timeout increase for now; but would also like to improve things so it's not actually slow00:47
jheskethif we're not repacking the servers, I think that might help00:48
jeblairjhesketh: i think we are, but it's worth verifying that is working correctly00:48
jheskethokay, I was going to look at that once the timeout is fixed as repacking only avoids the problem for so long00:48
SpamapSjeblair: at some point we should discuss how you'd like to handle a handoff from non-infra-root to infra-root for these servers.00:49
*** tjones1 has quit IRC00:49
SpamapSLike, do we need to produce reproduction instructions so you can build from scratch, or can we just leave our rootkits on it and give you the password? ;)00:49
*** miqui has quit IRC00:50
greghaynesreproduction instructions: sudo bash <(curl
*** tqtran has quit IRC00:51
greghaynesdon worry, I promise no rootkits00:51
SpamapSgreghaynes: genius00:51
SpamapSso this is really annoying00:51
SpamapSI was seeing DHCP reqs on brbm before00:52
SpamapSnow I"m not00:52
SpamapSand it takes 8 minutes to test any theories00:52
greghaynesSpamapS: on brbm? shouldnt brbm be all static?00:52
SpamapSgreghaynes: brbm is just bridged from the network00:52
SpamapSits bridged from em300:53
SpamapSnot wanting to unwind all of that :-P00:53
SpamapSgreghaynes: about to switch to virtual media. ;)00:53
*** ivar-lazzaro has quit IRC00:53
*** ivar-lazzaro has joined #openstack-infra00:54
greghaynesSpamapS: so the dhcp req's youre hoping to see, thats when youre telling your node to reboot pxe?00:54
*** ivar-lazzaro has quit IRC00:54
greghaynesid expect that to happen over em0 or em1 which wont be on the same l2 AIUI?00:54
SpamapSgreghaynes: I'm just pxe booting into the Ubuntu installer00:55
*** ivar-lazzaro has joined #openstack-infra00:55
SpamapSgreghaynes: em3 is the only interface up00:55
SpamapSgreghaynes: on bastion00:55
*** mtanino has quit IRC00:56
jheskethjeblair: okay, so did you want to merge that haproxy update? (I can test its effectiveness)00:57
*** yolanda has quit IRC00:57
SpamapSgreghaynes: so the DHCP reqs are coming in on brbm00:58
SpamapS17:57:37.166800 78:e7:d1:22:95:ed > ff:ff:ff:ff:ff:ff, ethertype IPv4 (0x0800), length 423: > BOOTP/DHCP, Request from 78:e7:d1:22:95:ed, length 38100:58
*** yolanda has joined #openstack-infra00:58
SpamapSbut dnsmasq is not responding00:58
greghaynesI see them00:58
SpamapSudp        0      0    *                           8624/dnsmasq00:58
SpamapSwhich is weird00:58
SpamapSawww man00:58
SpamapSthere's a bunch of crap in iptables00:59
*** jamielennox is now known as jamielennox|away01:00
*** abitha has quit IRC01:00
jeblairianw: please don't send automated mails to the infra list01:02
*** alexsyip has quit IRC01:02
SpamapSMonitor is in graphics mode or an unsupported text mode.01:04
jeblairjhesketh: i can't stick around to babysit it right now, sorry01:06
jeblairSpamapS: let's do talk; perhaps at next infra meetng?01:06
*** Longgeek has quit IRC01:06
*** jamielennox|away is now known as jamielennox01:08
SpamapSjeblair: sounds good. would that be a good topic for the infra ML too?01:08
*** yamamoto has joined #openstack-infra01:08
jeblairi think we can handle it at the meeting01:08
ianwjeblair: i expected people not to want that, i just sent one as an example of the output.  however, i think this needs monitoring, i hope i can find people who are interested other than me01:08
ianwat the moment, anyway, it just about fails more than it succeeds01:09
SpamapSjeblair: k01:10
jeblairianw: i approved the sentinal change, and jhesketh is working on the fundamental problem.  honestly though, i don't want anyone to have to care if an image doesn't update for a few days.01:12
ianwjeblair: well having just spent too much time debugging such a problem, i'd disagree.  when images end up with version skew, and some tests pass and others fail, it's a nightmare01:12
*** Daviey has quit IRC01:13
jeblairianw: i want to fix things so they go back to 'usually succeeding' instead of 'usually failing'.  and we've known they usually fail for a while now, it's just only recently that jhesketh has identified a root cause.01:13
openstackgerritMerged openstack-infra/nodepool: Enhance logging around image building
openstackgerritMerged openstack-infra/irc-meetings: Fix incorrect docs meeting time
jeblairianw: i mostly agree with you.  i think we need a nice status display of when each image was updated01:13
*** Daviey has joined #openstack-infra01:13
jeblairianw: that way, when there is a problem, it's easy to check that for a potential cause01:13
jeblairianw: the place i don't want to go is that we would have an expectation that every image would update every day, or that it would be the job of one or more person to check that and care01:14
jeblairianw: if we wanted to have nagios send an email after an image didn't update for 4 days in a row, i could get behind that.01:15
*** AK has quit IRC01:15
jheskethjeblair: no trouble re merging01:15
greghaynesI wonder if we did testing per-commit of image builds similar to the new dib tests if it would make the failures more obvious01:15
ianwgreghaynes: that is a good extra step, but doesn't help if the image didn't build01:16
jeblairgreghaynes: that would be good, but that's not why they are all failing now01:16
greghaynesyes, it definitely doesnt solve the problem01:16
jeblairianw: i believe greghaynes is proposing that we test changes to our build scripts to make sure we don't break the builds ourselves01:17
greghaynesyes, and also to make transient failures more obvious01:17
jeblairianw: (so we would test them by building an image using dib)01:17
jeblairgreghaynes: right, under the theory that if the external world is broken, it will show up in our tests01:18
fungisort of back now and caught up on recent scrollback. scaling the git cluster was my primary recommendation when discussing this with jhesketh earlier. pretty sure we're just asking the backends to do more than they have capacity for (because of nova and other large repos, and also the increasing number of third-party ci and other systems doing full clones at the same time ours do because they copy01:18
fungiour nodepool config, timer values included)01:18
openstackgerritMerged openstack-infra/irc-meetings: Adding Ops-tag team biweekly meeting
jeblairfungi: ugh.  perhaps we should make our config a secret value.01:18
jeblairer our times01:19
jeblairianw: anyway, a nodepool status page would be a huge benefit; i don't think we need a full rest api for that01:19
*** jyuso1 has joined #openstack-infra01:19
jeblairianw: also, when we finish the dib work, we will not have version skew on images01:21
openstackgerritMerged openstack-infra/irc-meetings: Create individual ICS files and add to webpage
jeblairthey will be identical01:21
*** puranamr has quit IRC01:21
*** Kennan has quit IRC01:22
*** Kennan has joined #openstack-infra01:23
fungiyeah, i envision the nodepool status page at least starting out like the zuul status interface... just a status.json we can transform with some javascript01:23
jeblairso i think the best places to focus our efforts are: 1) merge the timeout increase to mask the git problem; 2) nodepool dib; 3) nodepool status page; 4) scale git farm01:23
ianwwell, yes, but i'm talking about today.  the barrier to debugging failures is already incredibly high.  when those failures are happening on a homogeneous underlying environment it makes things even worse01:23
jeblairianw: the dib stuff is almost finished01:24
jeblairianw: wait, it's _worse_ when the failures are on a homogenous environment?01:25
jeblairianw: i thought version skew was a problem.  seems like a homogenous environment is exactly what we want?01:25
ianwoh i guess i mean heterogeneous ...01:25
ianwanyway, having one kernel oops and another not because it's got the right packages really sucks, that's my point01:26
* fungi was about to be confused01:26
jeblairyeah.  let's just get the dib work done asap.  i'm pretty sure it should be able to be done in a day or two if we can convince mordred to tell us what needs to be done, or have someone else take over it.01:26
ianwi'm happy to spend some time on that status.json thing.  i don't want to overlap with the full REST api thing, which i don't think i have cycles to work on, atm anyway01:27
jeblairianw: cool, i think we can find the review cycles for it too.01:27
mordredjeblair: I'm back in normal timezones - will work on finishing it off tomorrow01:27
greghaynesjeblair: I think were waiting on someone to test it with rax01:27
jeblairmordred: w00t, thx01:27
greghaynesmordred: ^ correct?01:28
mordredyah. that's my todo for when I wake up01:28
jeblairokay, i really have to run now, dinner's calling.  ttyl.01:28
lifelessfungi: yay - is green - but stalled until we get the devstack patch merged01:29
*** bhunter71 has quit IRC01:32
reazemHello :) How would one go about renaming a stackforge project?01:35
*** bhunter71 has joined #openstack-infra01:38
ianwlifeless: as someone who can +2 that devstack review, and is happy it passed ci, but hasn't followed closely so is about 80% sure it won't break the world ... can you double convince me it's safe? :)01:39
lifelessit safe01:39
ianw2015-06-17 23:23:45.791 | + echo '*** /dev/vdb appears to already be mounted'01:40
ianw2015-06-17 23:23:45.792 | *** /dev/vdb appears to already be mounted01:40
ianw^ wtf is up with the centos job now :(01:40
*** hdd has joined #openstack-infra01:45
fungiianw: booting with ephemeral disk premounted?01:45
*** asrangne has joined #openstack-infra01:46
ianw/dev/vdb        /mnt    auto    defaults,nofail,comment=cloudconfig     0       201:48
ianwmordred: any idea who might know why this has started happening?01:49
fungioh, that's the configdrive01:49
fungiwas that in rackspace?01:49
*** asrangne__ has quit IRC01:49
ianw/dev/vdb: LABEL="ephemeral0" UUID="043e49b1-9d5e-4120-aff5-8e1e8d90eb2e" TYPE="ext4"01:50
fungioh okay that is the ephemeral disk? any indication what size it was?01:51
*** asrangne has quit IRC01:51
*** ivar-lazzaro has quit IRC01:51
ianwfungi: <- 439G01:51
ianwi see it mounted when i just boot a centos host on hpcloud, though01:51
fungithe configdrive would be really tiny, so that's definitely the ephemeral disk01:51
ianwsomething has changed to make it start mounting it01:52
fungiwe normally mount it to /opt01:52
*** coreycb has quit IRC01:52
jheskethfungi: do you have time to baby sit per chance?01:52
jheskethno worries if not :-)01:52
fungii wonder whether glean thought that was the configdrive?01:52
fungijhesketh: nope, going to sleep in moments01:53
jheskethfair enough :-)01:53
fungiianw: are you seeing that when you boot a nodepool/dib/whatever created centos image, or just the hpcloud base image?01:53
ianwfungi: hpcloud base image has it mounted01:54
greghaynesI dont think glean has an error message like that01:54
fungiianw: oh, yeah that wouldn't be something we've done to it them01:54
ianwfungi: no, i guess the "comment=cloudconfig" in fstab points to the culprit01:55
fungiianw: oh, is cloud-init mounting that?01:55
fungiif not, i'm about out of ideas01:55
ianwif it is, it doesn't log about it :(01:56
greghaynesianw: where are you seeing this?01:56
ianwgreghaynes: boot a centos image in hpcloud, it will have /dev/vdb mounted01:57
ianwgreghaynes: ssh centos@ (i got your keys from github)01:58
fungii _think_ hpcloud no longer makes their own images and just hosts copies of distro-provided cloud images instead, so in theory this is a behavior intrinsic to official centos images01:59
*** coreycb has joined #openstack-infra02:00
ianwfungi: yeah, except i don't think the image has changed since it was last working02:00
fungiianw: also sounds likely. i think they use a new image name when they update images, so in theory we'd have had to change our nodepool config to get new base image behavior02:01
*** woodster_ has quit IRC02:01
greghaynesI think cloud-init adds that02:01
ianw /etc/cloud/cloud.cfg  has mounts module, so that makes sense.  what doesn't make sense is why this just started happening02:02
greghaynesianw: I wonder if its being served in the metadata now for some reason?02:03
fungithat would certainly be something that could alter boot-time behavior without altering the base image02:03
*** dkranz has quit IRC02:05
ianwso did hpcloud just decide to add that?02:05
fungidefinitely there now, so a plausible explanation02:05
greghaynesyea, no idea if thats a change or why, but that seems to be the culprit02:06
ianwwell, i guess i'll just work-around it in devstack-gate02:06
fungii guess stripping out cloud-init will solve that risk. short-term we can tweak our nodepool prep scripts to configure cloud-init to not automount that02:07
*** mtanino has joined #openstack-infra02:07
fungias an alternative02:07
greghaynesheh, now we have more ammo for that decision02:08
*** xyang0 has joined #openstack-infra02:10
*** koolhead17 has joined #openstack-infra02:11
*** xyang0 has quit IRC02:11
ianwdoes metadat override /etc/cloud/... values?02:12
*** maxandersen has quit IRC02:12
greghaynes/etc/cloud? I didnt know there was a data source for there02:14
ianw sorry i mean /etc/cloud/cloud.cfg.d/foo.cfg02:15
ianwwell, i don't know ... i dropped in a cfg file to override mounting in there, and it didn't mount02:15
ianwbut then i removed it, and it didn't mount it either :(02:16
greghaynesI dont think thats a datasource, I think you can configure datasources in there02:16
*** coreycb has quit IRC02:16
greghayneslooks like you might be able to pass datasource specific overrides or somesuch, not sure how that works02:16
ianwyeah, override "mounts:" in there02:16
ianwtell it to ignore ephemeral002:17
openstackgerritMerged openstack-infra/gearman-plugin: Use TextParameterValue instead of String
*** sabeen1 has joined #openstack-infra02:17
greghaynesI dont think that overrides the data in the datasource, I think that dir is just for configuring the datasource itself02:17
lifelessianw: sorry I was on the phone before02:18
lifelessianw: the devstack patch is what you were concerned about ?02:18
*** tjones1 has joined #openstack-infra02:19
ianwlifeless: yeah, the pip install of requirements02:19
lifelessianw: so yeah, its safe - the requirements repo is bog standard02:19
lifelessianw: there was a bug where it didn't install enough, but that doesn't affect devstack until we decide to move from -> update-requirements02:20
*** jamielennox is now known as jamielennox|away02:20
lifelessianw: which can be deferred indefinitely IMO02:20
mesteryEasy peasy infra patch for anyone who has a minute:
lifelessianw: the patch does two things; its moves a git clone after installing pbr; and it pip installs requirements02:21
anteayamestery: please let that be the title of the commit message02:21
lifelessianw: requirements has two non-conflicting entry point scripts (generate-constraints and update-requirements)02:21
lifelessianw: and its dependencies (nothing today, parsely in future)02:21
*** maxandersen has joined #openstack-infra02:21
*** maxandersen has joined #openstack-infra02:21
anteayaah I'm already +202:22
lifelessianw: and there is no dependency between the git clone and the pip install so the ordering there is irrelevant; but we want pbr present before running pip install $thingusingpbr02:22
lifelessianw: hope that helps you assess it02:23
ianwlifeless: ok, that makes sense; so it's mostly to get the generate-constraints/update-requirements?02:23
lifelessianw: its to unblock
mesteryanteaya: lol :)02:23
ianwgreghaynes: ok, it does do what i thought, you just gotta rm /var/lib/cloud/instances so cloud-init does its thing again02:24
lifelessianw: which was blocked by not having parsely installed in devstack - and as you can see its green with the devstack patch applied :)02:24
ianwlifeless: oh, i get it, get the parser in, which wasn't required before.  ok cool02:25
*** fallenpegasus has joined #openstack-infra02:26
lifelessianw: <3 thank you02:27
*** rfolco has quit IRC02:28
*** weshay has quit IRC02:30
*** unicell has quit IRC02:31
*** dannywilson has joined #openstack-infra02:31
*** dims has quit IRC02:32
*** dims has joined #openstack-infra02:33
ianw <- so we actually ignore the metadata02:33
ianwbut the *first* time it boots, cloud-init has made that fstab entry02:34
*** tjones1 has quit IRC02:35
*** dannywilson has quit IRC02:36
*** fallenpegasus has quit IRC02:37
*** AzherKhan has joined #openstack-infra02:37
*** ashleighfarnham has joined #openstack-infra02:37
*** coreycb has joined #openstack-infra02:38
*** Shrews has quit IRC02:39
*** Shrews has joined #openstack-infra02:39
*** aswadr has joined #openstack-infra02:40
*** dims has quit IRC02:41
*** fallenpegasus has joined #openstack-infra02:41
*** esmute has quit IRC02:42
*** ashleighfarnham has quit IRC02:42
openstackgerritDoug Wiegley proposed openstack-infra/project-config: Add lbaasv1 api job to the check queue, non-voting
*** Zhongjun has joined #openstack-infra02:43
*** otter768 has joined #openstack-infra02:45
*** Longgeek has joined #openstack-infra02:48
*** otter768 has quit IRC02:50
*** jyuso1 has quit IRC03:01
*** stevemar has joined #openstack-infra03:01
*** jyuso1 has joined #openstack-infra03:02
*** jamielennox|away is now known as jamielennox03:03
openstackgerritTan Lin proposed openstack/diskimage-builder: Add oat-client element
*** ramashri has joined #openstack-infra03:13
openstackgerritJohn L. Villalovos proposed openstack-infra/irc-meetings: Add href URL for the IRC channel in template
*** jamielennox is now known as jamielennox|away03:15
*** woodster_ has joined #openstack-infra03:18
*** enikanorov2 has quit IRC03:18
*** garyk has quit IRC03:19
*** ramashri has quit IRC03:22
*** Longgeek has quit IRC03:22
*** ramashri has joined #openstack-infra03:22
*** mriedem1 has quit IRC03:22
*** dannywilson has joined #openstack-infra03:23
*** jamielennox|away is now known as jamielennox03:25
*** tlian has quit IRC03:25
sdake_say any inra cats aroudn at this hour - question about multinode gating03:29
*** tjones1 has joined #openstack-infra03:40
*** koolhead17 has quit IRC03:45
*** koolhead17 has joined #openstack-infra03:46
*** ramashri has quit IRC03:48
*** enikanorov2 has joined #openstack-infra03:50
*** garyk has joined #openstack-infra03:51
*** yanfengxi has joined #openstack-infra03:51
*** unicell has joined #openstack-infra03:51
*** dannywilson has quit IRC03:52
*** dimtruck is now known as zz_dimtruck03:52
*** mtanino has quit IRC03:54
*** sdake_ has quit IRC03:56
*** unicell has quit IRC03:58
*** esmute has joined #openstack-infra04:00
*** harlowja_at_home has joined #openstack-infra04:01
*** unicell has joined #openstack-infra04:01
*** esmute has quit IRC04:01
*** esmute has joined #openstack-infra04:03
StevenKcinerama: Huh, I can't reproduce your zanata-cli bug04:05
StevenKcinerama: WRT not working with :443 in the URL04:05
openstackgerritIan Wienand proposed openstack-infra/project-config: Remove mount of ephemeral disk
openstackgerritIan Wienand proposed openstack-infra/project-config: Remove cloud-init details
*** achanda has joined #openstack-infra04:07
*** SergK has quit IRC04:07
*** pasquier-s has quit IRC04:07
*** esmute has quit IRC04:08
*** esmute has joined #openstack-infra04:10
*** tjones1 has quit IRC04:11
*** bhunter71 has quit IRC04:11
*** Longgeek has joined #openstack-infra04:19
*** unicell has quit IRC04:21
*** unicell has joined #openstack-infra04:21
*** ildikov has quit IRC04:22
*** achanda has quit IRC04:23
*** yfried_ has quit IRC04:24
*** unicell1 has joined #openstack-infra04:25
*** abitha has joined #openstack-infra04:25
*** unicell has quit IRC04:26
*** hdd has quit IRC04:26
*** abitha has quit IRC04:30
*** coolsvap|away is now known as coolsvap04:30
*** rossella_s has quit IRC04:30
*** rossella_s has joined #openstack-infra04:30
*** achanda has joined #openstack-infra04:33
jlvillaltonyb, On the ICS file patch to irc-meetings04:37
jlvillaltonyb, I see the webpage is updated, but the actual ICS file doesn't seem to be available.  For example:
*** yamada-h has joined #openstack-infra04:39
*** achanda_ has joined #openstack-infra04:39
*** achanda_ has quit IRC04:40
*** sdake_ has joined #openstack-infra04:41
*** enikanorov2 has quit IRC04:41
jlvillaltonyb, I do notice there is a 'meetings' directory already.  Maybe I picked a bad directory name :(
*** achanda has quit IRC04:41
*** boris-42 has joined #openstack-infra04:44
*** otter768 has joined #openstack-infra04:46
*** fedexo has joined #openstack-infra04:48
*** abregman has joined #openstack-infra04:49
*** clayton has quit IRC04:50
*** Longgeek has quit IRC04:51
*** otter768 has quit IRC04:51
*** vinay_yadhav has joined #openstack-infra04:55
*** puranamr has joined #openstack-infra04:59
*** dvorak has joined #openstack-infra05:00
*** yamada-h_ has joined #openstack-infra05:09
*** Kennan has quit IRC05:09
*** Kennan has joined #openstack-infra05:10
*** yamada-h has quit IRC05:11
*** wenlock has joined #openstack-infra05:13
*** yamada-h_ has quit IRC05:13
*** wenlock has quit IRC05:13
*** yfried_ has joined #openstack-infra05:15
*** xyang1 has quit IRC05:16
*** vinay_yadhav has quit IRC05:16
*** esmute has quit IRC05:19
*** deepakcs has joined #openstack-infra05:20
*** Swami has joined #openstack-infra05:23
*** jamielennox is now known as jamielennox|away05:23
*** dannywilson has joined #openstack-infra05:25
*** _nadya_ has joined #openstack-infra05:25
*** dannywilson has quit IRC05:25
*** dannywilson has joined #openstack-infra05:25
*** yfried_ is now known as yfried|afk05:26
*** jamielennox|away is now known as jamielennox05:26
*** coolsvap is now known as coolsvap|away05:28
*** hdd has joined #openstack-infra05:30
*** _nadya_ has quit IRC05:32
*** AzherKhan has quit IRC05:34
*** jyuso1 has quit IRC05:35
*** coolsvap|away is now known as coolsvap05:36
*** chlong has quit IRC05:38
*** yfried|afk is now known as yfried_05:41
*** sdake_ has quit IRC05:45
*** angela-s has joined #openstack-infra05:45
*** dannywilson has quit IRC05:48
*** Longgeek has joined #openstack-infra05:48
*** flepied has quit IRC05:49
*** sabeen1 has quit IRC05:49
*** chlong has joined #openstack-infra05:53
openstackgerritIan Wienand proposed openstack-infra/devstack-gate: Add option to setup netconsole to remote destination
*** jyuso1 has joined #openstack-infra05:57
*** mrunge has joined #openstack-infra05:57
*** Kennan has quit IRC05:58
*** Swami has quit IRC05:58
*** k4n0 has joined #openstack-infra05:58
*** enikanorov2 has joined #openstack-infra05:59
openstackgerritOpenStack Proposal Bot proposed openstack-infra/project-config: Normalize projects.yaml
*** Kennan has joined #openstack-infra06:01
*** fedexo has quit IRC06:02
*** AzherKhan has joined #openstack-infra06:04
*** koolhead17 has quit IRC06:04
*** ParsectiX has joined #openstack-infra06:05
*** Kennan has quit IRC06:06
*** Kennan has joined #openstack-infra06:06
*** harlowja_at_home has quit IRC06:09
*** dtantsur|sick is now known as dtantsur06:13
*** Longgeek has quit IRC06:14
*** hdd has quit IRC06:15
*** Zhongjun has quit IRC06:18
*** _nadya_ has joined #openstack-infra06:20
*** phiche has joined #openstack-infra06:22
*** yamada-h has joined #openstack-infra06:23
*** flepied has joined #openstack-infra06:25
*** flepied1 has joined #openstack-infra06:26
*** pasquier-s has joined #openstack-infra06:28
*** SergK has joined #openstack-infra06:29
*** flepied has quit IRC06:29
*** vinay_yadhav has joined #openstack-infra06:31
*** jyuso1 has quit IRC06:33
*** yamada-h_ has joined #openstack-infra06:34
*** kaisers has quit IRC06:37
*** yamada-h has quit IRC06:37
*** yamada-h has joined #openstack-infra06:38
*** puranamr has quit IRC06:41
*** yamada-h_ has quit IRC06:42
*** bhunter71 has joined #openstack-infra06:45
*** bhunter71 has quit IRC06:46
*** mpaolino has joined #openstack-infra06:46
*** otter768 has joined #openstack-infra06:47
*** kaisers has joined #openstack-infra06:47
*** Hal has joined #openstack-infra06:47
*** Hal is now known as Guest2355806:48
*** yolanda has quit IRC06:49
*** melwitt has quit IRC06:51
*** woodster_ has quit IRC06:51
*** otter768 has quit IRC06:52
*** armax has quit IRC06:53
*** coolsvap is now known as coolsvap|away06:53
*** ParsectiX has quit IRC06:54
*** kaisers has quit IRC06:54
*** melwitt has joined #openstack-infra06:57
*** ihrachyshka has joined #openstack-infra06:58
*** afazekas has joined #openstack-infra07:00
*** coolsvap|away is now known as coolsvap07:01
*** stevemar has quit IRC07:02
*** kaisers has joined #openstack-infra07:04
*** e0ne has joined #openstack-infra07:04
*** vinay_yadhav has quit IRC07:05
*** ildikov has joined #openstack-infra07:06
*** Ala has joined #openstack-infra07:06
*** Ala has quit IRC07:06
*** Ala has joined #openstack-infra07:08
*** yanfengxi has quit IRC07:09
*** jcoufal has joined #openstack-infra07:09
*** vaidy_ has joined #openstack-infra07:13
*** camunoz is now known as camunoz_gone07:14
*** Longgeek has joined #openstack-infra07:15
*** btully has quit IRC07:18
*** chlong has quit IRC07:18
*** fdegir has joined #openstack-infra07:19
*** btully has joined #openstack-infra07:19
*** BharatK has joined #openstack-infra07:20
*** vaidy_ has quit IRC07:20
*** e0ne has quit IRC07:23
*** jyuso1 has joined #openstack-infra07:26
*** ParsectiX has joined #openstack-infra07:27
*** devvesa has joined #openstack-infra07:28
*** arxcruz has joined #openstack-infra07:31
*** coolsvap is now known as coolsvap|away07:34
*** samuelBartel has joined #openstack-infra07:34
*** markus_z has joined #openstack-infra07:35
*** jlvillal has quit IRC07:37
*** [HeOS] has quit IRC07:38
*** salv-orl_ has quit IRC07:40
*** jlanoux has joined #openstack-infra07:40
*** dizquierdo has joined #openstack-infra07:48
*** boris-42 has quit IRC07:52
*** SotK has joined #openstack-infra07:53
*** jlvillal has joined #openstack-infra07:54
*** dizquierdo has quit IRC07:55
*** joeljwright has joined #openstack-infra07:55
*** yfried_ is now known as yfried|afk07:56
*** yfried|afk is now known as yfried_07:57
*** joeljwright has quit IRC07:57
*** shardy has joined #openstack-infra07:57
openstackgerritLuis Cañas-Díaz proposed openstack-infra/activity-board: Rename 'Tickets Participants' with 'Launchpad Participants'
*** dizquierdo has joined #openstack-infra08:01
*** abregman has quit IRC08:01
*** dizquierdo has quit IRC08:01
*** joeljwright has joined #openstack-infra08:02
*** joeljwright has quit IRC08:02
*** ZZelle has quit IRC08:03
*** ZZelle has joined #openstack-infra08:03
*** phiche1 has joined #openstack-infra08:03
*** ruagair has quit IRC08:03
*** hashar has joined #openstack-infra08:03
*** Longgeek has quit IRC08:04
*** Longgeek has joined #openstack-infra08:04
*** abregman has joined #openstack-infra08:05
*** liam_ has joined #openstack-infra08:05
*** liam_ has quit IRC08:05
*** phiche has quit IRC08:05
*** bexelbie has joined #openstack-infra08:07
*** ihrachyshka_ has joined #openstack-infra08:08
openstackgerrityolanda.robla proposed openstack-infra/nodepool: Allow to delete nodes with missing target or provider
*** ihrachyshka has quit IRC08:09
*** cdent has joined #openstack-infra08:11
openstackgerritMerged openstack-infra/irc-meetings: Add href URL for the IRC channel in template
*** dguerri` is now known as dguerri08:15
openstackgerritMerged openstack-infra/release-tools: Add
openstackgerritMerged openstack-infra/release-tools: let the caller specify the project owner
openstackgerritMerged openstack-infra/release-tools: use announce list instead of dev list
openstackgerritMerged openstack-infra/release-tools: Add
*** kushal has joined #openstack-infra08:18
openstackgerritThierry Carrez proposed openstack-infra/release-tools: Add script
openstackgerritDavide Guerri proposed openstack-infra/shade: Centralize exception management for Neutron
openstackgerritDavide Guerri proposed openstack-infra/shade: Add Neutron/Nova Floating IP list/search/get
*** hashar has quit IRC08:23
*** dougshelley66 has quit IRC08:23
*** dougshelley66 has joined #openstack-infra08:24
*** abregman has quit IRC08:28
*** ihrachyshka_ has quit IRC08:28
*** vinayyadhav has joined #openstack-infra08:30
*** vinay_yadhav has joined #openstack-infra08:30
*** rossella_s has quit IRC08:30
*** rossella_s has joined #openstack-infra08:31
*** vinayyadhav has quit IRC08:32
*** vinay_yadhav has quit IRC08:32
*** vinayyadhav has joined #openstack-infra08:32
*** vinay_yadhav has joined #openstack-infra08:32
*** notnownikki has joined #openstack-infra08:35
*** ParsectiX has quit IRC08:36
*** joeljwright has joined #openstack-infra08:37
*** derekh has joined #openstack-infra08:40
*** abregman has joined #openstack-infra08:40
*** tsufiev has quit IRC08:40
*** ihrachyshka_ has joined #openstack-infra08:41
*** jtomasek has joined #openstack-infra08:41
*** tsufiev has joined #openstack-infra08:41
*** hashar has joined #openstack-infra08:42
*** pbourke has quit IRC08:44
*** dims has joined #openstack-infra08:44
*** pbourke has joined #openstack-infra08:45
*** e0ne has joined #openstack-infra08:48
*** otter768 has joined #openstack-infra08:48
*** dims has quit IRC08:48
*** vinay_yadhav has quit IRC08:49
*** vinayyadhav has quit IRC08:50
*** [HeOS] has joined #openstack-infra08:51
*** otter768 has quit IRC08:53
*** notnownikki has quit IRC08:56
*** isviridov_away is now known as isviridov08:56
*** e0ne is now known as e0ne_08:58
*** notnownikki has joined #openstack-infra08:58
*** Nikkau has joined #openstack-infra08:59
*** pelix has joined #openstack-infra09:00
*** ociuhandu has joined #openstack-infra09:00
*** e0ne_ is now known as e0ne09:00
*** dizquierdo has joined #openstack-infra09:00
*** Ala has quit IRC09:01
*** coolsvap|away is now known as coolsvap09:02
openstackgerritBradley Jones proposed openstack-infra/project-config: Create Magnum UI repo
*** Ala has joined #openstack-infra09:04
*** oomichi has quit IRC09:05
*** laughterwym has quit IRC09:05
*** dtantsur is now known as dtantsur|brb09:05
*** u_glide has left #openstack-infra09:07
*** spredzy|afk is now known as spredzy09:07
*** romcheg has joined #openstack-infra09:09
openstackgerritAntoine Musso proposed openstack/diskimage-builder: dib-lint: make it work on Mac OS X
openstackgerritAntoine Musso proposed openstack/diskimage-builder: dib-lint: make it work on Mac OS X
*** ihrachyshka_ is now known as ihrachyshka09:13
*** matbu_ has joined #openstack-infra09:14
*** ParsectiX has joined #openstack-infra09:20
*** yfried_ is now known as yfried|afk09:20
*** dizquierdo has quit IRC09:20
*** cgoncalves has joined #openstack-infra09:21
*** dizquierdo has joined #openstack-infra09:22
cgoncalvesHi. I must be missing something obvious, but it's the second time I rebase my branch and git-review and still Jenkins does not verify because of a merge conflict:
cgoncalvesI successfully rebased in patch set 3 and now in patch set 409:23
*** matbu_ has quit IRC09:23
*** pbourke has quit IRC09:25
*** vinayyadhav has joined #openstack-infra09:29
*** vinay_yadhav has joined #openstack-infra09:29
electricalcgoncalves: could it be tht due to the 2 Depends-On dependencies it breaks merging?09:30
cgoncalveselectrical: mmm not sure if Jenkins takes that into consideration, good point.09:31
*** pbourke has joined #openstack-infra09:32
electricalits just a guess at this point. I barely worked with gerrit and not sure if it processes the dependencies as part of your change09:32
*** claudiub has joined #openstack-infra09:32
*** rguillebert has joined #openstack-infra09:32
*** joeljwright has quit IRC09:34
*** Guest23558 has quit IRC09:35
*** kiall has quit IRC09:35
*** Kiall has joined #openstack-infra09:36
*** Hal has joined #openstack-infra09:36
*** Hal is now known as Guest2773009:37
*** joeljwright has joined #openstack-infra09:38
*** yfried|afk is now known as yfried_09:40
*** ociuhandu_ has joined #openstack-infra09:46
*** ociuhandu has quit IRC09:47
*** ociuhandu_ is now known as ociuhandu09:47
*** kir_ has joined #openstack-infra09:47
kir_Hi all, today I re-install rally from scratch and got File "/home/kshileev/RVE/local/lib/python2.7/site-packages/rally/benchmark/", line 35, in <module>     from rally.plugins.openstack.context import existing_users as existingusers_ctx   File "/home/kshileev/RVE/local/lib/python2.7/site-packages/rally/plugins/openstack/context/", line 30, in <module>09:48
kir_Is it known problem? Any workaround?09:48
kir_sorry wrong channel please ignore09:49
*** yamamoto has quit IRC09:49
*** kir_ has left #openstack-infra09:50
*** yamamoto has joined #openstack-infra09:50
*** jasondotstar has joined #openstack-infra09:52
*** e0ne is now known as e0ne_09:57
*** e0ne_ is now known as e0ne09:59
*** Guest27730 has quit IRC10:00
*** dims has joined #openstack-infra10:00
*** yfried_ is now known as yfried|afk10:01
*** jcoufal has quit IRC10:02
*** vinay_yadhav has quit IRC10:05
*** vinayyadhav has quit IRC10:05
*** yamamoto has quit IRC10:07
*** dtantsur|brb is now known as dtantsur10:08
ihrachyshkaflaper87, any ideas why grenade for neutron fails on glance error (NotAuthenticated: Authentication required)?
openstackgerritDarragh Bailey proposed openstack-infra/jenkins-job-builder: Test update command with --delete-old option
*** yamamoto has joined #openstack-infra10:11
electricalfungi: running into some issues with jjb. most likely i'm trying to abuse it again :p10:12
pelixany JJB or infra core reviewers around to review ?10:13
electricalhiya pelix10:13
electricalpelix: {es_versions_{es_branch}} <-- any idea if this is possible? haha.   so i have an array of branches in es_branch that i create the jobs for based on job template. and i have variables called es_versions_master and es_versions_1x  that have specific details for that branch10:14
*** Longgeek_ has joined #openstack-infra10:20
*** achanda has joined #openstack-infra10:20
*** yfried|afk is now known as yfried_10:21
*** Longgeek has quit IRC10:23
*** dizquierdo has quit IRC10:23
pelixelectrical: hmm, I don't think there it is possible for python's 'format()' to correctly interpolate inner and then outer10:25
*** samuelBartel has quit IRC10:26
*** salv-orlando has joined #openstack-infra10:26
electricalpelix: i was afraid of that :-(10:26
openstackgerritAntoine Musso proposed openstack/diskimage-builder: doc: migrate from README.rst to Sphinx
pelixPossible you could try multiple macros?10:26
*** samuelBartel has joined #openstack-infra10:26
electricalpelix: macros in what way? the values are a list for the matrix jobs. not sure how to use that with a macro?10:28
*** salv-orl_ has joined #openstack-infra10:29
*** hashar has quit IRC10:30
pelixelectrical:, read the section on 'Format Specifiers' which points out that you cannot have replacement fields nested to arbitrary levels.10:30
*** salv-orlando has quit IRC10:31
*** yfried_ is now known as yfried|afk10:31
sdaguelifeless: so - got merged, but I'm not really comfortable with that long term. If requirements has to be installed for the rest of this to work it seems like it needs to be a library or a dependency.10:31
sdagueotherwise, requirements of requirements can mask things that have to be right with projects10:32
*** joeljwright has quit IRC10:33
pelixelectrical: define multiple macros for the scm section one for each of the branches and then just use es_versions_{es_branch} to pick up the desired macro10:34
KiallAny infra folks around who can approve a project creation? should be good to go + governance change landed a few days back10:34
*** mdenny has quit IRC10:34
electricalpelix: in this case its not really a branch in scm but a list of items in the matrix options.  ( axis parts )10:35
*** mdenny has joined #openstack-infra10:35
electricalwould that work as well to define a macro for axis?10:36
pelixelectrical: does that mean that what you want is for the template to produce ${es_versions_1x} in the resulting job?10:36
pelixmight be easier if you paste something to a gist10:36
electricalwill do that.10:36
*** maishsk_afk has joined #openstack-infra10:36
*** Hal has joined #openstack-infra10:39
*** mdenny has quit IRC10:39
*** kmartin has quit IRC10:39
*** Hal is now known as Guest6494110:39
*** mdenny has joined #openstack-infra10:39
*** kmartin has joined #openstack-infra10:39
*** joeljwright has joined #openstack-infra10:39
*** yfried|afk is now known as yfried_10:40
*** achanda has quit IRC10:41
electricalpelix: hope this makes a bit more sense10:42
*** shardy_ has joined #openstack-infra10:42
*** mdenny has quit IRC10:43
*** kmartin has quit IRC10:43
*** mdenny has joined #openstack-infra10:44
*** shardy has quit IRC10:44
*** kmartin has joined #openstack-infra10:44
*** hichihara has quit IRC10:46
*** mpaolino has quit IRC10:47
*** shardy_ has quit IRC10:48
*** shardy has joined #openstack-infra10:49
*** otter768 has joined #openstack-infra10:49
*** yfried_ is now known as yfried|afk10:50
*** e0ne is now known as e0ne_10:50
openstackgerritDmitry Tantsur proposed openstack/requirements: Add python-ironic-inspector-client to projects.txt
dtantsurcould someone please review this ^^^ and the parent patch? thanks!10:51
*** vinay_yadhav has joined #openstack-infra10:52
*** vinayyadhav has joined #openstack-infra10:52
*** jlanoux has quit IRC10:53
*** otter768 has quit IRC10:54
flaper87ihrachyshka: looking10:55
*** coolsvap is now known as coolsvap|away10:56
*** ildikov_ has joined #openstack-infra10:56
*** yfried|afk is now known as yfried_10:56
flaper87ihrachyshka: probably caused by that10:57
flaper87not sure what's going on with glance+keystone there10:57
flaper87I wonder if it has something to do with all the keystone v3 changes10:58
ihrachyshkaflaper87, is it httpd dying?10:58
flaper87ihrachyshka: i don't think it died because it returned 50010:58
* flaper87 checks keyston's logs10:58
*** ildikov has quit IRC10:58
flaper87wait, aren't we collecting httpd's logs ?11:00
ihrachyshkaflaper87, from oslo_utils import excutils11:00
ihrachyshkaflaper87, they are collected11:00
flaper87ah nvm11:00
*** e0ne_ has quit IRC11:00
ihrachyshkaok, so it's not glance, sorry for hassle :)11:00
*** btully has quit IRC11:00
flaper87ihrachyshka: no worries, glad I helped11:01
*** btully has joined #openstack-infra11:02
*** marcusvrn has joined #openstack-infra11:02
*** AK has joined #openstack-infra11:05
pelixelectrical: I think I have a way of solving this11:05
*** joeljwright has quit IRC11:06
*** AzherKhan has quit IRC11:07
pelixelectrical: checkout my comment on
*** mwagner_afk has quit IRC11:10
*** AK has quit IRC11:11
ihrachyshkasdague, can you look at that keystone upgrade issue? ^^ seems like we (sometimes?) catch import failure in keystone in grenade11:12
ihrachyshkacan it be related to eg.
*** matbu_ has joined #openstack-infra11:12
sdagueihrachyshka: that's because they are copying the wsgi script instead of using an installed entry point11:12
sdagueso it's not getting upgraded11:12
ihrachyshkasdague, do we have a fix for that?11:13
sdagueI don't know of anything posted11:13
sdaguethat was part of my reason for the wsgi thread11:14
ihrachyshkaok, going to the openstack-dev@ to read11:14
*** yfried_ is now known as yfried|afk11:15
*** matbu_ has left #openstack-infra11:15
*** yfried|afk is now known as yfried_11:15
openstackgerritMartin Pavlásek proposed openstack-infra/project-config: Enable horizon integration tests voting
*** asrangne has joined #openstack-infra11:19
*** asrangne__ has joined #openstack-infra11:20
*** aswadr has quit IRC11:21
*** devvesa has quit IRC11:21
ihrachyshkasdague, what makes you think that the copied script is the culprit? It seems that what really crashes is somewhere deep in code located in proper place. apparently, at that same moment when keystone was trying to start, we get this: "Successfully uninstalled oslo.utils-1.6.0". we should not update libs while keystone is supposed to run; if anything, a restart is needed after we settled. thoughts?11:23
*** asrangne has quit IRC11:24
openstackgerritMerged openstack-infra/project-config: Make heatclient-dsvm-functional gating for heatcli
dtantsurhi infra! i was assuming that pushing a signed tag to gerrit will make a release, if publish-to-pypi jobs are used. was my assumption wrong? project in question is python-ironic-inspector-client11:25
ihrachyshkaah wait, there is another failure later in the same log, now with "ArgsAlreadyParsedError: arguments already parsed: cannot register CLI option" That one could be relevant to copied script indeed, depending on its contents. where do they get the file?11:25
openstackgerritMartin Pavlásek proposed openstack-infra/project-config: Enable horizon integration tests voting
dtantsur(I would check myself, but does not show anything in "release" queue)11:26
*** sdake has joined #openstack-infra11:26
openstackgerritMartin Pavlásek proposed openstack-infra/project-config: Enable horizon integration tests voting
openstackgerritMartin Pavlásek proposed openstack-infra/project-config: Enable horizon integration tests voting
*** jlvillal has quit IRC11:30
mordreddtantsur: ok, I agree that there is not a package there - looking deeper11:30
dtantsurgit os-job gives "file not found" :(11:31
dtantsurmordred, thanks!11:31
*** jlvillal has joined #openstack-infra11:31
*** hashar has joined #openstack-infra11:32
mordred2015-06-18 11:13:18.737 | + tox -evenv python sdist11:33
mordred2015-06-18 11:13:18.861 | ERROR: unknown environment 'venv'11:33
*** amotoki_ has joined #openstack-infra11:33
mordreddtantsur: you don't have a venv venv in your repo11:33
dtantsurmordred, ghmm, it wasn't mentioned anywhere as a dependency...11:33
dtantsurmordred, so I should create a dummy one to make it happy, right?11:34
mordrednope. you should make a proper one that works11:34
mordredone sec - patch coming11:34
*** pasquier-s has quit IRC11:35
*** pasquier-s has joined #openstack-infra11:35
dtantsurmordred, yeah, I understand, it will just be identical to py27, I guess... thanks anyway!11:36
*** e0ne has joined #openstack-infra11:36
mordreddtantsur: land that and tag 1.0.1 and you should be good!11:36
mordreddtantsur: it's the envionment we use to run arbitrary commands - like build_sphinx or sdist11:37
dtantsurmordred, awesome! how hard is the requirement to have "cover" and sphinx support? will something break if I don't have them for the time being?11:37
electricalpelix: just got back from lunch. I see the solution yeah. would have loved to have it in a single thing but this will work as well. thanks!11:37
mordreddtantsur: I mean, you should have them in general, but nothing will break if you don't have the jobs at the moment11:38
mordreddtantsur: so no immediate rush - but at some point it shoudl make it on to someone's todo list11:38
dtantsurmordred, understood, thanks again! (btw I'm used to have coverage report as part of usual test run - thus no seperate -ecover)11:38
*** pc_m has joined #openstack-infra11:38
mordreddtantsur: yah- that one is the least important - I think being set up for docs is more important11:39
*** yfried_ has quit IRC11:41
*** jasondotstar has quit IRC11:46
openstackgerritMartin Pavlásek proposed openstack-infra/project-config: Enable horizon integration tests voting
*** dalgaaf has joined #openstack-infra11:49
jheskethmordred: oh hiya11:54
jheskethmordred: any chance you feel like helping babysit
*** btully has quit IRC11:54
*** baoli has joined #openstack-infra11:54
*** ldnunes has joined #openstack-infra11:55
*** kushal has quit IRC11:55
*** btully has joined #openstack-infra11:55
*** vinay_yadhav has quit IRC11:57
*** vinayyadhav has quit IRC11:58
*** dizquierdo has joined #openstack-infra11:58
*** markus_z has quit IRC11:58
*** mpaolino has joined #openstack-infra11:59
electricalpelix: seems it doesn't work when the es_versions_1x and es_versions_master are in an other defaults namespace then global :(12:00
*** joeljwright has joined #openstack-infra12:00
*** pblaho has quit IRC12:01
electricalI might have to reconsider the design i had in mind.12:01
electricalBecause i wanted to have separate name spaces per projects12:02
*** achanda has joined #openstack-infra12:02
*** dpaterson has joined #openstack-infra12:03
*** deepakcs has quit IRC12:05
*** yamada-h has quit IRC12:07
*** tosky has joined #openstack-infra12:07
*** pblaho has joined #openstack-infra12:08
toskyhi all, if you have some time for two patches which fixes horizon integration tests, it would be nice!12:09
tosky reenable Sahara for horizon integration tests, fixing the last error; and makes them voting again12:09
*** dprince has joined #openstack-infra12:09
*** kushal has joined #openstack-infra12:10
*** jasondotstar has joined #openstack-infra12:10
*** salv-orl_ has quit IRC12:11
*** Guest64941 has quit IRC12:14
*** rfolco has joined #openstack-infra12:14
*** Shrews has quit IRC12:17
mordredjhesketh: sure!12:17
*** Nikkau has quit IRC12:18
*** Shrews has joined #openstack-infra12:18
openstackgerritMerged openstack-infra/activity-board: Rename 'Tickets Participants' with 'Launchpad Participants'
*** kushal has quit IRC12:18
*** bknudson has joined #openstack-infra12:20
*** btully has quit IRC12:22
*** weshay has joined #openstack-infra12:23
*** yamamoto has quit IRC12:23
openstackgerritJulia Kreger proposed openstack/diskimage-builder: Move simple-init data files out of phase subdirs
*** btully has joined #openstack-infra12:25
*** kgiusti has joined #openstack-infra12:25
*** isviridov is now known as isviridov_away12:25
*** bknudson has quit IRC12:26
*** BharatK has quit IRC12:27
*** puranamr has joined #openstack-infra12:28
*** ajmiller has quit IRC12:28
*** aysyanne has joined #openstack-infra12:29
*** rossella_s has quit IRC12:30
*** mfmcdonagh has quit IRC12:30
*** mcdonagm has quit IRC12:30
*** rossella_s has joined #openstack-infra12:30
*** tlian has joined #openstack-infra12:32
*** e0ne is now known as e0ne_12:32
*** jlanoux has joined #openstack-infra12:33
*** puranamr has quit IRC12:35
*** dguitarbite has quit IRC12:38
*** e0ne_ is now known as e0ne12:39
*** dims has quit IRC12:40
*** dizquierdo is now known as dizquierdo_afk12:41
*** dims has joined #openstack-infra12:41
*** yfried_ has joined #openstack-infra12:42
*** jasondotstar has quit IRC12:42
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Don't sanitize node names for log directories
*** bknudson has joined #openstack-infra12:43
*** yolanda has joined #openstack-infra12:43
openstackgerritMerged openstack-infra/project-config: Enable Sahara for Horizon integration tests
*** woodster_ has joined #openstack-infra12:45
mordredjeblair, Shrews, SpamapS: I am currently testing the shade swift patch against rackspace12:46
Shrewsmordred: wunderbar12:47
*** prad has joined #openstack-infra12:48
ihrachyshkasdague, to me, it does not seem like the old copied file since that file was not modified since Aug 201412:49
*** otter768 has joined #openstack-infra12:50
*** bswartz has quit IRC12:50
ihrachyshkasdague, also, if you can, please raise the prio for , I don't have perms12:51
openstackLaunchpad bug 1466485 in Keystone "keystone fails with: ArgsAlreadyParsedError: arguments already parsed: cannot register CLI option" [Undecided,New]12:51
sdagueihrachyshka: I'm working through zeromq plugin for devstack right now, so not really available to debug12:51
ihrachyshkasdague, ack, thanks anyway.12:52
*** eharney has quit IRC12:53
*** puranamr has joined #openstack-infra12:53
*** jasondotstar has joined #openstack-infra12:55
*** otter768 has quit IRC12:55
*** mohankumar has joined #openstack-infra12:55
mohankumarHI Team12:55
mohankumarI am trying to contribute, but get the following error:12:56
mohankumar"git review12:56
mohankumarfatal: ICLA contributor agreement requires current contact information.12:56
*** bexelbie has quit IRC12:56
mohankumarVing same email address in Openstack and girrit account12:57
mohankumarcould any one help me to resolve this issue12:57
*** bexelbie has joined #openstack-infra12:58
EmilienMmordred: I don't see progress on getting a gem mirror ( ) - do you know if there is any blocker?12:58
mordredEmilienM: I don't think so - I think it just hasnt hit people's review queues13:00
EmilienMmordred: no problem, I was just wondering :)13:02
openstackgerritMonty Taylor proposed openstack-infra/shade: Pass token and endpoint to swift os_options
mordredShrews, SpamapS: ^^13:02
mordredjhesketh: ^^ you might like that too13:02
*** julim has joined #openstack-infra13:03
AJaegermohankumar: Did you follow the instructions from Otherwise check and13:06
*** salv-orlando has joined #openstack-infra13:07
jheskethmordred: lgtm13:07
*** doug-fish has joined #openstack-infra13:08
mordreddguerri: you may be interested in teh above as well13:09
*** dustins has joined #openstack-infra13:09
electricalmordred: AJaeger Morning gents13:10
AJaegermorning, electrical !13:10
electricalFew more months and I'm off to Portland for a week ;-)13:11
mohankumarThanks AJaeger  for the pointer !!13:11
*** achanda has quit IRC13:12
*** salv-orl_ has joined #openstack-infra13:12
electricalAnyone here going to PuppetConf this year?13:12
openstackgerritGorka Eguileor proposed openstack-infra/elastic-recheck: Update query for bug 1412513
openstackbug 1412513 in Cinder "Intermittent unit test failures for incorrect notifier count" [Undecided,Confirmed] - Assigned to Eric Harney (eharney)13:13
*** salv-orlando has quit IRC13:15
pelixelectrical: seems to work for me when in order defaults13:15
pelixdid you remember to name the which defaults where to be used in the job-template?13:16
electricalpelix: yeah i got the same defaults name in the job-template13:16
openstackgerritDmitry Tantsur proposed openstack/requirements: Add python-ironic-inspector-client to global-requirements
openstackgerritJohn L. Villalovos proposed openstack-infra/irc-meetings: Change name of calendar dir to 'calendars'
pelixelectrical: looks like it's due to a bug fixed on master but not in a released version yet13:17
electricalpelix: ah. i'm running of a branch which includes the defaults merging thing13:18
electricalSince i needed that feature13:18
*** jasondotstar has quit IRC13:18
pelixodd, should be fixed for you so13:19
*** puranamr has quit IRC13:19
electricalHmm, time to rebase the master branch and re-patch it :-)13:20
mordredShrews: ok- the image seems to have uploaded to swift at least, so that's a good start13:21
mordredit's attempting to import into glance now13:21
mordredSpamapS: Manager rax ran task <shade._tasks.ObjectCreate object at 0x7f21adaf2210> in 0.00208497047424s13:21
mordredSpamapS: it's not urgent, but I'm pretty sure that's not a correct time :)13:22
*** rguillebert has quit IRC13:22
electricalpelix: that one could use some more eyes13:22
openstackgerritFausto Marzi proposed openstack-infra/project-config: Splitting stackforge/freezer on multiple repos
*** yamamoto has joined #openstack-infra13:24
*** ayoung is now known as ayoung-mtg13:25
*** bswartz has joined #openstack-infra13:27
*** bcourt has joined #openstack-infra13:27
*** mriedem has joined #openstack-infra13:27
Shrewsmordred: w00t13:28
mordredShrews: dont woot too much - the import failed13:28
*** ujuc has joined #openstack-infra13:29
*** yamamoto has quit IRC13:29
*** dkranz has joined #openstack-infra13:30
mordrednikhil_k: I don't suppose there is a way to know _why_ an image task-import failed if the error code is 396 and the message is "Image cannot be imported. Error code: '396'" is there?13:30
*** ashleighfarnham has joined #openstack-infra13:30
*** dboik has quit IRC13:30
openstackgerritKun Huang proposed openstack-infra/irc-meetings: Add tricircle meeting
pelixelectrical: looks like it could do with a minor fix to drop the support of the old 'room' entry since it would require a migration to publisher of the yaml to activate in the first place.13:32
*** signed8bit has joined #openstack-infra13:34
*** fallenpegasus has quit IRC13:34
*** ashleighfarnham has quit IRC13:35
*** fifieldt_ is now known as fifieldt13:36
*** rguillebert has joined #openstack-infra13:36
*** mwagner_afk has joined #openstack-infra13:37
*** hashar_ has joined #openstack-infra13:37
*** peristeri has joined #openstack-infra13:38
*** hashar has quit IRC13:38
EmilienMpuppet group is motivated by making vote puppet4 jobs: - if anyone can have a look :-)13:39
*** dizquierdo_afk is now known as dizquierdo13:40
*** puranamr has joined #openstack-infra13:40
*** mnestratov has joined #openstack-infra13:41
mordredEmilienM: awesome!13:41
*** dguitarbite has joined #openstack-infra13:41
EmilienMmordred: I'm curious to see if I don't break your modules CI13:42
EmilienMI'm checking now13:42
mnestratovHi all13:42
mnestratovJust found out pretty funny fact about CI comments. If CI reports "Build succeeded (check pipeline)" or "Build failed (check pipeline)"13:42
mnestratovthis report is put in the first row in CI reports table rewriting existing first entry. So did our CI and when it reported after13:42
mnestratovJenkins, it rewrote its table entry, otherwise Jenkins after our CI rewrote ours. Though we have just fixed this, I wonder is this a well known fact?13:42
mnestratovExamples of Jenkins missing table entries in review can be found here,n,z13:42
*** ayoung-mtg is now known as ayoung13:42
mordredEmilienM: I don't think we've added any pupept4 jobs to our things yet, have we?13:42
EmilienMno indeed13:42
EmilienMyou don't run rspec13:43
EmilienMso w're good I'll say13:43
electricalpelix: good point.13:43
mordredShrews, SpamapS, nikhil_k: ok. I have downloaded the image from swift and verified that the thign that's in swift is identical to the vhd image I have locally13:44
*** dannywilson has joined #openstack-infra13:44
*** dannywilson has quit IRC13:44
*** bexelbie has quit IRC13:44
mordredso the swift upload using swift service part seems to be operating properly13:44
*** amitgandhinz has joined #openstack-infra13:44
*** bexelbie has joined #openstack-infra13:45
pelixelectrical: I'm not opposed to keeping it but then it should log a warning that it's a deprecated setting13:45
mordredand I ran the import again from the command line using glance and it worked13:45
*** dannywilson has joined #openstack-infra13:45
*** puranamr has quit IRC13:45
mordredlet me delete taht image and try again - maybe it was just an oops13:45
electricalpelix: yeah indeed. But it makes sense to remove it since its basically a new plugin13:45
*** vinay_yadhav has joined #openstack-infra13:46
*** eharney has joined #openstack-infra13:46
*** vinayyadhav has joined #openstack-infra13:46
EmilienMinfra is not kidding when you ask for a patch here. They merge it in less than 5 minutes. (thx mordred & fungi)13:46
*** puranamr has joined #openstack-infra13:46
mordredEmilienM: we aim to please!13:46
*** garyk has quit IRC13:46
*** tvanderwiel has joined #openstack-infra13:48
fungiEmilienM: well, i saw you ping about it yesterday but needed to wait for a moment when i could spot check to make sure your branch matching was valid for your repos (since you used to have series branch names with no stable/ prefix)13:50
*** dannywilson has quit IRC13:50
*** vinay_yadhav has quit IRC13:50
*** vinayyadhav has quit IRC13:50
EmilienMfungi: cool. I'll make sure nothing is broken on our side but I think we're good13:51
openstackgerritRoman Podoliaka proposed openstack-infra/project-config: Run oslo.db unit tests with MySQL-Python too
*** vinayyadhav has joined #openstack-infra13:51
*** vinay_yadhav has joined #openstack-infra13:51
openstackgerritMartin Pavlásek proposed openstack-infra/project-config: Enable horizon integration tests voting
*** achanda has joined #openstack-infra13:52
*** dboik has joined #openstack-infra13:52
nikhil_kmordred: hey13:53
jheskethI'm off for the night, but if somebody is around to baby sit that'd be great :-)13:53
nikhil_kI am about to go for a series of meetings (to chair) so, will try to answer that question13:53
nikhil_kerror 396 I am not sure what that is. It may be one of the scurity checks failed though intermittently13:54
mordrednikhil_k: cool - it _seems_ like it may have just been a temporary error - but I'd love to know whether it's something I can respond to in code13:54
mordrednikhil_k: kk13:54
marzif_Hi Infra, please review at your convenience : )
nikhil_kmordred: gotcha, yeah. it may have been an intermittent conn error at any of the inter service integration points13:54
nikhil_kor auth token failure or something like that13:54
nikhil_kerror messages are grouped to identify the type of failure13:55
nikhil_kand support knows them thoroughly, though there amy be a wiki but I don't know for sure13:55
*** mcdonagm has joined #openstack-infra13:55
*** mfmcdonagh has joined #openstack-infra13:55
nikhil_kbut I can try to help :) if stuck13:56
mordrednikhil_k: awesome- I appreciate it - I'm retrying a few times to see if it happens often or if I just got lucky13:57
*** sigmavirus24_awa is now known as sigmavirus2413:57
*** jasondotstar has joined #openstack-infra13:58
openstackgerritMonty Taylor proposed openstack-infra/shade: Update segmented object metadata after creation
*** jasondotstar has quit IRC13:58
*** hdd has joined #openstack-infra13:59
*** bexelbie_ has joined #openstack-infra13:59
*** maishsk_afk is now known as maishsk14:01
*** ParsectiX has quit IRC14:02
*** achanda has quit IRC14:02
*** achanda has joined #openstack-infra14:03
*** achanda has quit IRC14:03
*** EricGonczer_ has joined #openstack-infra14:04
*** yamamoto has joined #openstack-infra14:04
mordredSpamapS: oh! the debug time is off because of the threads and whatnot14:04
openstackgerritMerged openstack-infra/elastic-recheck: Update query for bug 1413034
openstackbug 1413034 in OpenStack-Gate "Download of failed" [Undecided,Confirmed]
*** abregman has quit IRC14:05
*** ddieterly has joined #openstack-infra14:06
openstackgerritMerged openstack-infra/project-config: puppet: make Puppet 4 unit test jobs voting
openstackgerritJohn Griffith proposed openstack-infra/elastic-recheck: Add query for bug 1465831
openstackbug 1465831 in Cinder "testtools.matchers._impl.MismatchError: 0 != 15 running test_create_delete_snapshot" [Undecided,Invalid]
*** puranamr has quit IRC14:08
*** yamamoto has quit IRC14:08
*** k4n0 has quit IRC14:08
*** btully has quit IRC14:09
*** afazekas has quit IRC14:09
*** kushal has joined #openstack-infra14:10
*** ddieterly has quit IRC14:10
*** btully has joined #openstack-infra14:10
*** yamahata has joined #openstack-infra14:12
*** ddieterly has joined #openstack-infra14:13
*** stevemar has joined #openstack-infra14:13
*** nfedotov has joined #openstack-infra14:14
*** bhunter71 has joined #openstack-infra14:14
*** btully has quit IRC14:15
*** oomichi has joined #openstack-infra14:15
*** puranamr has joined #openstack-infra14:16
*** oomichi has quit IRC14:16
*** ddieterly has quit IRC14:16
*** btully has joined #openstack-infra14:17
*** hashar_ is now known as hashar14:17
openstackgerritValeriy Ponomaryov proposed openstack-infra/project-config: Fix Manila jobs for running on stable/juno branch
*** yamamoto has joined #openstack-infra14:19
openstackgerritSteve Kowalik proposed openstack-infra/project-config: Support pushing translations to Zanata
*** thedodd has joined #openstack-infra14:23
openstackgerritValeriy Ponomaryov proposed openstack-infra/project-config: Fix Manila jobs for running on stable/juno branch
*** unicell has joined #openstack-infra14:24
*** btully has quit IRC14:24
*** unicell1 has quit IRC14:25
*** btully has joined #openstack-infra14:25
timrc_"Warning: Found multiple default providers for vcsrepo: hg, cvs, bzr, svn, git; using hg" <--- Huh? hg? Puppet, why you so strange?14:26
*** xyang0 has joined #openstack-infra14:26
mordredtimrc_: you know ... because ...14:27
mordredtimrc_: sorry, I have zero good reasons14:27
mordredI believe that the world needs to admit that git is now the default, and while the others are perfectly valid alternatives that people can love and care for14:28
*** hdd has quit IRC14:28
mordredthe default VCS provider for no tool should, at this point, be anything that is not git14:28
timrc_We're cloning from git so why Puppet says its going to use hg... that's beyond me :)14:28
openstackgerritMerged openstack-infra/shade: Pass token and endpoint to swift os_options
mordredtimrc_: I blame nibalizer14:28
timrc_It's obviously not using hg so maybe I'm misunderstanding that statement :)14:28
*** BharatK has joined #openstack-infra14:29
timrc_mordred, I do that on a daily basis :)14:29
mordredtimrc_: I believe that is happening at class instantiation time in the vcsrepo provider14:29
mordredtimrc_: basically, it's letting you know that it has found multiple drivers that all claim to be default ,so it's picking one14:29
mordredtimrc_: but this is happening _before_ it reads any of your parameters14:30
timrc_mordred, Ah, so just another way to make puppet a little slower, needlessly.14:30
*** yamahata has quit IRC14:30
*** hdd has joined #openstack-infra14:30
*** mpaolino has quit IRC14:30
mordredtimrc_: it's important to find as many of those ways as possible and make use of them14:31
timrc_The I should blame vcsrepo specifically and not Puppet generally :)14:31
timrc_<-- unfair14:31
*** jasondotstar has joined #openstack-infra14:31
mordredvcsrepo is a dirty wet dog - it's just the best one we have14:31
fungior more to the point, blame our ~3-years-stale fork of the vcsrepo module14:31
timrc_It's a sound strategy.  Introduce a bunch of inefficiencies, acquire a large user base, strategically remove those inefficiencies, prosper.14:31
*** btully has quit IRC14:32
*** jasondot_ has joined #openstack-infra14:32
sdaguejeblair / fungi - what is the next step on the ci-watch side? I uploaded a spec - - but I'm not sure where we go from here (process wise), because it's more of a concept plan14:32
mordredtimrc_: ++14:32
sdagueI guess I should mordred ^^^ that as well14:32
mordredfungi: of the things our non-fork will fix, I don't think that is one of them14:32
fungimordred: yep14:33
*** tonytan4ever has joined #openstack-infra14:33
*** ajmiller has joined #openstack-infra14:33
*** zz_dimtruck is now known as dimtruck14:33
fungisdague: i still need to find time to read it, sorry :/14:33
mordredsdague: yah - same here14:33
sdaguefungi: ok, no worries, I just don't want to have things stall out because folks like tchaypo wanted to start helping on it as well14:34
*** btully has joined #openstack-infra14:34
openstackgerritMerged openstack-infra/shade: Convert ironicclient node.validate() call to Task
jeblairsdague: let's try to revise the spec next week, and get it approved the week after.  i'm sprinting on the team guide the rest of this week, and have a couple specs ahead of it in my queue, so can't work on it right away.14:36
sdaguejeblair: ok14:36
*** jonaspf has quit IRC14:36
*** e0ne is now known as e0ne_14:37
*** e0ne_ is now known as e0ne14:37
*** puranamr has quit IRC14:38
sdaguejeblair: well, I'm gone in 2 weeks time. So we should front load this into next week if at all possible.14:39
openstackgerritRuslan Kamaldinov proposed openstack-infra/project-config: Make yaql py34 job voting
*** miqui has joined #openstack-infra14:40
jeblairsdague: ack, i think we can focus on the bulk of the work next week.14:41
*** fallenpegasus has joined #openstack-infra14:41
sdaguejeblair: great14:42
openstackgerritMerged openstack-infra/shade: Return True/False for delete methods
*** jasondotstar has quit IRC14:42
*** ihrachyshka has quit IRC14:43
yolandajeblair, what's expected zuul behaviour when you revert the parent change of a depends-on? should it be reverting only that parent , or all the childs that depended from it?14:43
*** Ala has quit IRC14:44
*** coreycb has quit IRC14:46
jeblairyolanda: it's undefined; a depends on just depends on a single patch.  whatever happens beyond that is out of zuul's control.14:47
*** oomichi has joined #openstack-infra14:47
*** romcheg has quit IRC14:48
yolandajeblair, so reversal should be done by hand? i'm just testing some zuul upgrades14:48
jeblairyolanda: yes14:48
yolandaok thx14:48
*** mtanino has joined #openstack-infra14:49
jeblairi'm going to be working on the project team guide sprint today, so i may not be very responsive.14:49
*** jasondotstar has joined #openstack-infra14:49
*** jasondotstar has quit IRC14:50
*** zz_jgrimm is now known as jgrimm14:51
*** jasondot_ has quit IRC14:51
*** romcheg has joined #openstack-infra14:51
*** otter768 has joined #openstack-infra14:51
*** jasondotstar has joined #openstack-infra14:51
*** whitenoise has quit IRC14:51
mordredAND - now all of the shade patches are in merge conflict :)14:52
mordreddguerri: sorry bout that14:53
yolandahere, we are testing replication on gerrit 2.8, and i find it's not creating some repos. It's normally creating all missing ones on remote mirrors, but missing a few. No errors on log14:53
dguerrino worries14:53
yolandaanyone had that problem?14:53
mordredyolanda: we don't use gerrit to create remote repos at all14:53
yolandamordred, how do you do?14:53
mordredyolanda: we use jeepyb create-cgit-repos on the cgit mirrors14:53
yolandawhy don't you use it? don't trust it?14:53
mordredyolanda: this is why we started using ansible to manage our puppet runs14:54
*** cbx33 has joined #openstack-infra14:54
mordredyolanda: because we need to run puppet on the cgit mirrors before on gerrit14:54
cbx33Hi all, having issues updating contact information14:54
*** mdbooth has joined #openstack-infra14:54
yolandamordred, don't tell us we need to move to ansible to properly replicate our mirrors :)14:54
eantyshevjeblair: hello, may i ask you to remove -1 on
mordredyolanda: well, you should be moving to ansible to run puppet ANYWAY because it's more resilient14:54
cbx33I just get "cannot store contact details"14:54
cbx33when i hit save14:54
sdaguepleia2: - seems like a thing that would be great for you to weigh in on14:55
jeblaireantyshev: did you do anything to address my comments?14:55
Shrewsmordred: so, um... wow. nice conflict magic you've got going on there14:55
mordredyolanda: and, I mean, also, it seems like gerrit may not be as good at creating repos as you'd like it to be14:55
mordredShrews: right?14:55
mordredShrews: I'm MAGICAL14:55
yolandamordred, it isn't14:55
*** otter768 has quit IRC14:55
Shrewsmordred: like a pony14:55
mordredyolanda: I blame java14:55
yolandait's skipping a few without any good reason14:56
mordredShrews: I certainly bite people like I'm a pony14:56
*** spzala has joined #openstack-infra14:56
mordredyolanda: hunch - something failed the first time, but it has the local repo, so it does not bother trying to create the remote one, because it thinks it's good14:56
eantyshevjeblair: I created unittest as a separate review
yolandaand yes, we'll love to move to ansible, but too many things to do....14:56
*** yamahata has joined #openstack-infra14:56
yolandaso it creates the local repo, and the local replica into /var/lib/git14:57
jheskethBy the way, who here is going to Oscon and would it be good to do a mini meetup around that?14:57
mordredyolanda: I hear that - well, maybe once we're done moving to puppet apply the combination of ansible+puppet-apply will be compelling enough for you to take the cost14:57
yolandabut for some reason gerrit will know, it doesn't create the remote ones14:57
mordredyolanda: yah - never used that gerrit feature, so I'm very little use14:57
yolandawe'll like to embrace it, and get rid to puppetmaster14:57
Clintjhesketh: i'm a maybe14:57
openstackgerritMonty Taylor proposed openstack-infra/shade: Update segmented object metadata after creation
openstackgerritMonty Taylor proposed openstack-infra/shade: Ensure that service values are strings
jeblaireantyshev: ah great, thanks!  i'll take a look at it soon14:58
mordredjhesketh: I will not be going to oscon14:58
*** puranamr has joined #openstack-infra14:58
mordredjhesketh: I will be on vacation in panama14:58
jheskethLucky you :-)14:59
mordredjhesketh: I promise to enjoy myself14:59
fungijhesketh: i'll be there. have a talk to give on thursday morning14:59
cbx33Code Review - Error15:00
cbx33Server Error15:00
cbx33Cannot store contact information15:00
cbx33That's the full error I get from review.openstack.org15:00
*** maishsk has quit IRC15:00
*** Hal has joined #openstack-infra15:00
*** jonaspf has joined #openstack-infra15:01
*** Hal is now known as Guest7949715:01
*** MarkAtwood has joined #openstack-infra15:02
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet: add 3.8 release tests support
*** puranamr has quit IRC15:03
EmilienMjust rebased ^15:03
*** achanda has joined #openstack-infra15:04
*** e0ne has quit IRC15:04
mordreduploading a 1.8G file to the cloud over and over again is ... boring15:04
sdagueso, I realize folks are sprinting, but this would be really nice to have - - especially as jogo is soon to go poof15:04
mordrednikhil_k: I reran the import and it worked - so it seems like that was just a temp error15:04
electricalmordred: really? i always found it interesting to see how fast it would go :p lol15:05
*** dangers_away is now known as dangers15:05
mordredelectrical: hah!15:05
*** phiche1 has left #openstack-infra15:05
sdaguebasically just executes the devstack worlddump if things go south in a devstack-gate run15:05
electricalmordred: and try to get it going faster the next time :p lol15:05
sdaguewhich gives us a pretty detailed post mortem15:05
mordredsdague: looks great15:05
*** Guest79497 has quit IRC15:06
*** bexelbie has quit IRC15:06
*** dboik has quit IRC15:06
joeljwrightHey everyone, really hoping for some help with some gate-tempest-dsvm-neutron-src-python-swiftclient errors I'm seeing15:07
cbx33Does anyone know who deals with issues on
*** dboik has joined #openstack-infra15:07
joeljwrightlots of tests failing with {u'message': u'No valid host was found. There are not enough hosts available.', u'code': 500, u'created': u'2015-06-18T14:42:29Z'}15:07
fungicbx33: lots of us in here. what instructions were you following? hopefully ?15:07
EmilienMcbx33: what kind of issue?15:07
joeljwrightcan anyone provide some insight?15:07
cbx33fungi, EmilienM I have my account setup, but I can't save new contact information15:07
fungicbx33: you have your account set up according to that documentation, or you simply think you have your account set up but haven't followed that documentation to do it?15:08
openstackgerritAlvaro del Castillo proposed openstack-infra/activity-board: Added search field feature.
EmilienMcbx33: your Fullname + prefered email would be enough15:08
fungicbx33: if you've gone back through the documentation and nothing there provides you with a resolution, there are some suggestions at of things you can double-check15:09
*** sabeen has joined #openstack-infra15:09
*** vinayyadhav has quit IRC15:09
*** vinay_yadhav has quit IRC15:10
*** boris-42 has joined #openstack-infra15:10
*** vinayyadhav has joined #openstack-infra15:10
*** vinay_yadhav has joined #openstack-infra15:10
cbx33EmilienM, so I tried pushing up and it told me the Individual Contributor thing needed contact information15:11
cbx33so i tried going back in to save them15:11
cbx33and i just get a black screen15:11
fungicbx33: unfortunately gerrit provides no means to configure that error message, and submitting your contact information is a systems cross-check to make sure you're set up with a corresponding foundation membership, so there are several potential reasons you might get that error (most of which imply that the documentation was not followed incompletely or incorrectly)15:11
cbx33saying the error I posted above15:11
yolandamordred, creating the repo with git init --bare in the mirror solved the issue and started replication, i wonder why gerrit keeps ignoring the creation of that project15:11
cbx33fungi, ok15:11
cbx33I will go back through it15:11
*** arxcruz has quit IRC15:11
fungier, imply that the documentation was followed incompletely or incorrectly15:12
fungicbx33: mostly you need to make sure that you have a foundation individual membership account for and that your primary/preferred e-mail address matches between that and gerrit15:12
SpamapSmordred: can you explain why that update is needed better?15:13
SpamapSmordred: I'd expect SwiftService to make sure the object is correct.15:13
mordredSpamapS: well, without the update the metadata is not set on the object15:13
*** achanda has quit IRC15:13
mordredSpamapS: which I determined by using shade to upload an image and then looking at swift state for the object15:13
mordredSpamapS: I agree with you - I would totally expect swiftservice to take care of that15:14
*** dannywilson has joined #openstack-infra15:14
fungijoeljwright: sounds like a resource problem in devstack when tested by tempest (possibly due to your proposed change). you might have better luck working through that with the devstack and tempest maintainers in #openstack-qa15:14
mordredSpamapS: hwever, I've learned to inhabit a plane of constant existential dread as it relates to things I expect openstack client libraries to do15:14
joeljwrightwell, I asked in openstack-qa and no-one can see a reason for it15:15
SpamapSmordred: I'd also expect a clue here:
joeljwrightso I was pointed here15:15
*** bexelbie_ has quit IRC15:16
mordredSpamapS: for the record- I did try setting those values in a "meta" dict in the options dict15:16
mordredSpamapS: and that also did not work15:16
*** dimtruck is now known as zz_dimtruck15:16
fungiyolanda: part of why we precreate those repos on our mirrors is that gerrit 2.4 did not support new repo creation over ssh and we implemented our remote mirrors while we were still running 2.4 and haven't revisited it since15:17
joeljwrightfungi: the proposed change leaves the code functionally identical (just renaming some non-kwarg method arguments)15:17
*** _nadya_ has quit IRC15:17
*** dannywil_ has joined #openstack-infra15:17
*** dannywilson has quit IRC15:17
joeljwrightfungi: so I'm totally at a loss15:17
*** sabeen3 has joined #openstack-infra15:17
mordredfungi: yup. although it sounds like we would not be pleased with the results15:17
yolandafungi, we are doing the test with gerrit 2.8, but it's really weird. I can see the task in gerrit queue, and suddenly it disappears without really creating the repo15:17
yolandano errors anywhere15:17
yolandatailing the secure log in centos i can see the connection and disconnection15:18
*** rwsu has joined #openstack-infra15:18
yolandabut doing nothing15:18
yolandaeven, no errors on centos messages15:18
*** sabeen has quit IRC15:18
fungijoeljwright: you asked when there weren't a lot of people awake in -qa. you might try again15:18
*** _nadya_ has joined #openstack-infra15:18
joeljwrightfungi: ok, thanks :) will go back15:19
fungijoeljwright: it's errors getting raised from tempest calls within a devstack environment, so there's not a lot we're going to be able to tell you from the infra side of things (we maintain the machines those jobs run on, but the tests themselves are more of a black box to us)15:19
*** patrickeast has joined #openstack-infra15:19
joeljwrightfungi: it's hard to work out when the various channels are most active sometimes15:19
*** oomichi has quit IRC15:19
mordredfungi, jeblair, SpamapS, Shrews: WOOT. image upload to rax completed (With those patches)15:20
mordredverifying I can boot it real quick15:20
joeljwrightfungi: great - at least I can go back with "not an infra issue"15:20
joeljwrightfungi: thanks for looking15:20
*** oomichi has joined #openstack-infra15:20
fungijoeljwright: well, not an infra issue until you can bring us infrastrucure-specific errors to investigate15:21
*** coreycb has joined #openstack-infra15:21
fungiyolanda: i think new repo creation over ssh was fairly new in 2.8... maybe it doesn't work well until later gerrit releases?15:22
*** oomichi has quit IRC15:22
*** oomichi has joined #openstack-infra15:22
yolandafungi, looks like, we'll need to adopt that remote creation using jeepyb15:22
*** sdake has quit IRC15:23
openstackgerritMerged openstack-infra/project-config: Create Magnum UI repo
mordredyolanda: how were you doing it with 2.4?15:24
mordredgreghaynes, SpamapS: danger will robinson15:24
yolandawe have a monitor to alert of that, and create manually15:25
mordredyolanda: oh. gross. yeah - just use create-cgit-repos15:25
fungiyolanda: though, my memory may be wrong. i'm finding references as far back as 2.0 that say "If replication is enabled, this command also connects to each of the configured remote systems over SSH and uses command line git on the remote system to create the empty repository."15:25
mordredyolanda: it's already puppet-worthy and everything15:25
*** oomichi has quit IRC15:25
fungibut i definitely remember something about repo creation over ssh in more recent release notes15:25
yolandafungi, it does, for most of them15:25
*** joeljwright has left #openstack-infra15:25
yolandabut it's missing a few on the way15:25
*** whitenoise has joined #openstack-infra15:25
openstackgerritMerged openstack-infra/project-config: Add a non-voting py34 check job to Cinder
openstackgerritMerged openstack-infra/project-config: Allow push merge commit for oslo.cache
yolandaso yes, we'll take a look at that tool15:26
openstackgerritMerged openstack-infra/project-config: Remove double curly brackets from cue builder
mordredSpamapS, greghaynes: image uploaded to rax worked. instance booted from image. I can't ping it15:26
mordredSpamapS, greghaynes: SO - I think we may have grown an issue with simple-init element15:26
SpamapSmordred: >:|15:26
*** devvesa has joined #openstack-infra15:26
*** armax has joined #openstack-infra15:26
mordredSpamapS, greghaynes: either of you have brainspace and/or rax accounts to help debug? (and then if it is a simple-init thing, we should figure out how to test for it in dib)15:27
SpamapSmordred: you said you tried with a 'meta' thing, the docs for SwiftService.upload say options['meta'] and options['headers'] should work.15:27
mordredthat said - I may be runnign old dib elements here - so I'm going to pull fresh and start from scratch15:27
mordredSpamapS: yes - I agree that the docs say that15:27
mordredSpamapS: I believe we may want to file a bug with them15:27
fungiyolanda: ahh, from the 2.8 release notes
fungi"The replication plugin now automatically creates missing repositories on the destination if during the replication of a ref the target repository is found to be missing."15:28
fungiso clearly in 2.8 they were still ironing this out15:28
yolandait does... sometimes15:28
SpamapSmordred: I may have found it actually15:29
SpamapSmordred: I believe that is supposed to be options['headers']15:30
*** cody-somerville has quit IRC15:31
*** hashar has quit IRC15:32
Shrewsyay, more client bugs in both doc and code15:32
*** gyee has joined #openstack-infra15:32
*** tonytan4ever has quit IRC15:33
openstackgerritSomay Jain proposed openstack-infra/jenkins-job-builder: Support for more extensions and options in Git plugin
SpamapSthe doc is right, the code is bong ;)15:33
*** tonytan4ever has joined #openstack-infra15:34
SpamapSjust digging through the tests now, there's no tests that headers ever make it through15:34
*** hodos has joined #openstack-infra15:34
SpamapSoh wow, there's almost no unit tests for SwiftService.upload15:35
*** sdake has joined #openstack-infra15:35
*** ramashri has joined #openstack-infra15:36
mordredSpamapS: yay!15:37
SpamapSswiftclient doesn't have coverage report turned on in their check jobs either15:37
* SpamapS waits for his box to compile lxml.. AGAIN15:37
*** rguillebert has quit IRC15:37
SpamapSreally need to figure out how to get tox to use a ccache or something15:37
fungiSpamapS: wheelhouse?15:38
fungipip should be creating a wheel of that and caching it15:38
*** Hal has joined #openstack-infra15:38
*** jogo has joined #openstack-infra15:39
SpamapSfungi: actually it did15:39
SpamapSI am so happy... I wasl ike "Hey, my tox is done already?!"15:39
*** achanda has joined #openstack-infra15:39
*** Hal is now known as Guest4445715:39
*** vinay_yadhav has quit IRC15:39
fungimad props to dstufft and crew15:39
*** vinayyadhav has quit IRC15:40
* mordred hands dstufft a mostly clean goose15:40
*** mdbooth has quit IRC15:40
Kiallmordred / fungi - Got 5 mins for a new repo review? hoping to land today so we don't have to re-do the git-foo to extract history again :)15:40
*** achanda_ has joined #openstack-infra15:41
openstackgerritRuslan Kamaldinov proposed openstack/requirements: Add yaql
*** afazekas has joined #openstack-infra15:42
*** achanda has quit IRC15:44
*** mdbooth has joined #openstack-infra15:45
mordredKiall: AJaeger beat me to it - but done15:45
KiallCool :)15:45
*** dtantsur is now known as dtantsur|afk15:46
*** BharatK has quit IRC15:48
*** unicell has quit IRC15:49
openstackgerritRuslan Kamaldinov proposed openstack-infra/project-config: Make yaql py34 job voting
*** rguillebert has joined #openstack-infra15:49
* SpamapS goes cross eyed looking for 'header' and 'headers' and "header" and "headers" in swiftclient's code15:50
*** afazekas has quit IRC15:50
*** achanda_ has quit IRC15:51
SpamapSit looks like they've been used interchangeably even in the documentation15:52
SpamapSthis is _amazing_15:52
*** achanda has joined #openstack-infra15:52
*** zz_dimtruck is now known as dimtruck15:52
*** patrickeast has quit IRC15:53
*** abregman has joined #openstack-infra15:53
*** dimtruck is now known as zz_dimtruck15:53
dstufftfungi: mordred Wheel caching was all lifeless really :) I just reviewed/merged15:53
openstackgerritDerek Higgins proposed openstack-infra/system-config: Remove tripleo-f20 nodes from tripleo cloud
openstackgerritDerek Higgins proposed openstack-infra/system-config: Set suitable min-ready values for tripleo fedora nodes
fungioh! i didn't realize he had hacked that part together. kudos lifeless!15:55
*** eharney has quit IRC15:55
*** zz_dimtruck is now known as dimtruck15:55
openstackgerritMonty Taylor proposed openstack-infra/shade: Update segmented object metadata after creation
* mordred hands lifeless a mostly clean slightly used goose15:57
mordredSpamapS, Shrews: ^^15:57
*** shardy_ has joined #openstack-infra15:58
*** ramashri has quit IRC15:59
openstackgerritAlvaro del Castillo proposed openstack-infra/activity-board: Added search field feature.
*** shardy has quit IRC16:00
*** _nadya_ has quit IRC16:00
*** enikanorov2 has quit IRC16:00
*** jasondotstar is now known as jasondotstar|afk16:01
SpamapSmordred: doc bug16:01
SpamapSmordred: the right option is 'header' not 'headers'16:01
openstackgerritMerged openstack-infra/devstack-gate: Run worldump if tempest fails
*** afazekas has joined #openstack-infra16:01
SpamapSmordred: lets get this into the functional test16:01
*** jasondotstar has joined #openstack-infra16:03
openstackgerritMerged openstack-infra/activity-board: Added search field feature.
*** jlanoux has quit IRC16:03
*** shardy_ has quit IRC16:04
mordredSpamapS: ++16:04
*** abregman has quit IRC16:04
*** shardy has joined #openstack-infra16:04
*** abregman has joined #openstack-infra16:05
mordredSpamapS: if we do it as header, it should work without the update?16:05
SpamapSmordred: correct16:05
SpamapSmordred: testing now16:05
mordredSpamapS: want me to test that locally with?16:05
mordredwith rackspace?16:05
SpamapSmordred: yes please!16:05
SpamapSmordred: literally one letter delete in the code16:05
SpamapSI just want it in our test :)16:05
*** nibalizer has quit IRC16:06
openstackgerritSean M. Collins proposed openstack-infra/project-config: Set some more defaults for Linux Bridge
anteayasc68cal: yay16:07
*** afazekas has quit IRC16:07
mordredSpamapS: btw is what I'm using to test16:08
SpamapSmordred: <-- aaaand its fixed in trunk. ;)16:08
mordredSpamapS: sweet16:09
*** rkukura has quit IRC16:09
*** bswartz has quit IRC16:10
mordredsc68cal: so - I'm guessing I could set up a shade functional test with those parameters and I'd have a neutron-based devstack to test against that does linuxbridge with direct attached ips?16:11
mordredsc68cal: as in - I could expect that cloud to not require me to use floating ips to get an externally routable ip?16:11
mordredSpamapS, Shrews: ^^ this is likely something we should set up, so that we can exercise the non-floating-ip side of our logic16:12
anteayamordred: this test is still experiemental16:12
anteayamordred: sc68cal is still finding and fixing bugs with it to get the right configuration16:12
anteayamordred: as well as get the right pieces for neutron in neutron and for devstack in devstack16:12
*** pbourke has quit IRC16:13
mordredthat's awesome- am I correct in my guess as to the intent of that particular test once it's up and going?16:13
anteayamordred: but concieveable once the work is done and the things are where they should be there would be a configuration for others to follow16:13
*** pbourke has joined #openstack-infra16:13
anteayaI don't know about the direct attached ip part16:13
anteayabut the point is to have linuxbridge working with devstack16:14
mordredeveryone in this channel is a giant fan of those16:14
mordredand I believe this is the thing that gets us that setup16:14
anteayaI know16:14
anteayaI don't know enough to confirm or dispute16:14
* mordred provisionally hands sc68cal a non-used donkey16:14
anteayaas I know so little about it16:14
anteayajust that pieces are in play16:14
anteayamordred: oh have you had a chance to look at the give me a network spec in neutron?16:15
anteayait got air time at yesterdays' neutron drivers meeting16:15
anteayaand will be discussed further at next week's sprint16:15
mordredanteaya: I have not had a chance to look at it16:15
* anteaya finds the link16:15
mordredanteaya: I support all of the parts of it that I would like, and I disagree with all the parts I wouldn't16:15
sigmavirus24does the openstackstatus bot not change topics back after it's replaced it with a gerrit notification?16:16
nikhil_kmordred: glad to hear it worked16:16
*** ihrachyshka has joined #openstack-infra16:16
mordrednikhil_k: \o/16:16
jeblairsigmavirus24: we fonud a bug in it last run; sorry; you can change it back manually now16:16
anteayamordred: great let's write them down so we don't have to guess next week16:16
mordrednikhil_k: so - non-urgent feedback ... "389" is not a particularly descriptive user-facing error :)16:16
*** ayoung is now known as ayoung-afk16:16
*** samuelBartel has quit IRC16:16
mordrednikhil_k: but, at this point, I'm not worried about it myself16:16
nikhil_kmordred: yeah16:16
anteayajaypipes: spec review reminder, give me a network:
sigmavirus24jeblair: no worries, me was looking for a bug to fix =P16:17
nikhil_kmordred: we had a bit fight product vs dev vs qe vs sucurity vs mgmt on that16:17
jeblairmordred:  are you _sure_? :)16:17
*** alexsyip has joined #openstack-infra16:17
jaypipesanteaya: yep, it's on my list.16:18
jeblairsigmavirus24: we haven't _fixed_ the bug in statusbot if you want to :)16:18
nikhil_kmordred: the tradeoffs were obfuscation (security centric) vs. user friendly error (support centric)16:18
anteayajaypipes: thank you16:18
sigmavirus24jeblair: is there a bug report for it or no?16:18
nikhil_kpeople just decided better safe*16:18
*** bdemers has quit IRC16:19
*** ujuc has quit IRC16:19
greghaynesmordred: did you figure out what the not getting networking issue in rax is?16:19
*** afazekas has joined #openstack-infra16:20
openstackgerritClint 'SpamapS' Byrum proposed openstack-infra/shade: Fix MD5 headers regression
jeblairsigmavirus24: there is now!!/story/200029816:22
SpamapSmordred: ^^ supersedes your patch16:22
anteayamordred: feel like reviewing and potentially giving a +3 so he can continue to figure out how to linuxbride with devstack?16:22
*** openstackgerrit has quit IRC16:22
greghaynesah, fun16:23
anteayahe being sc68cal16:23
*** openstackgerrit has joined #openstack-infra16:23
sc68calmordred: when you say directly attached IP, can you give a little more info?16:25
marzif_any traction is appreciated on
*** jasondotstar has quit IRC16:25
* sc68cal sees scrollback16:26
sc68calmordred: we'd have to document how to do provider networking in devstack with LB - we currently only have OVS documented16:26
*** jogo has quit IRC16:26
sc68cal(since I wrote it)16:26
sdaguesc68cal: sounds like you have a new todo :)16:27
*** romcheg has quit IRC16:28
sc68calsdague: oh I have to document using LB in devstack first - so TODO >= 216:28
mordredsc68cal, anteaya: commented on spec16:29
* sc68cal makes cards in trello16:29
anteayamordred: thank you16:30
*** unicell has joined #openstack-infra16:30
sc68calmordred: thanks :)16:30
mordredsc68cal: so - what I mean is that there are two different ways you can wind up with an externally routable ip on your VM16:30
anteayacan we give a shove too?16:30
*** rossella_s has quit IRC16:30
mordredsc68cal: you can have an IP from a private network allocated to you and you can get a floating ip to get out16:30
anteayamore config for linuxbridge with devstack16:30
mordredsc68cal: or you can get an IP from an externally routable network allocated to you directly16:30
*** afazekas has quit IRC16:30
*** rossella_s has joined #openstack-infra16:31
sc68calmordred: right - and that can be done with the provider networking API extension for Neutron16:31
mordredsc68cal: when you say api extension - is that an API extension that nova uses when talking to neutron? or a client-facing API?16:31
* mordred learns things ... is excited16:32
*** sputnik13 has joined #openstack-infra16:32
*** sputnik13 has left #openstack-infra16:32
sc68calmordred: So it's an API extension that the cloud admin uses to create a Neutron network and subnet - after creation it just looks like a regular neutron network16:32
mordredsc68cal: in your devstack patches, is the resulting devstack configured such that nova boot will get an ip on the externally-routable network without requiring a floating IP?16:32
*** jasondotstar|afk is now known as jasondotstar16:33
mordreds/devstack patches/devstack-gate patches/16:33
sc68calmordred: DevStack supports doing this kind of topology - although I only have documented it for OVS16:33
sc68calmordred: not at the gate16:33
mordredok. so that would still be some work we'd need to do if we wanted a devstack configured that way to do functional tests of shade on16:33
*** eharney has joined #openstack-infra16:34
sc68calmordred: yes, but it shouldn't require much, the config knobs are already present16:34
*** unicell has quit IRC16:34
mordredsc68cal: NEAT!16:34
mordredand thanks for those docs - they make sense to my brain16:35
sc68calI just need to update some parts of that doc by using the correct RFC space for IPv4 documentation16:35
*** unicell has joined #openstack-infra16:35
*** clu_ has joined #openstack-infra16:35
*** spzala has quit IRC16:35
sc68calusing was dumb16:35
sc68calI should use
sc68calthat'll make it more clear that the instances will get public, routed IPv4 addresses - without floating IPs16:36
*** devvesa has quit IRC16:36
*** vinayyadhav has joined #openstack-infra16:36
mordredSpamapS: ^^ yay!16:36
*** vinay_yadhav has joined #openstack-infra16:37
*** ryanpetrello has quit IRC16:37
anteayasc68cal: nice work16:37
*** samueldmq has quit IRC16:37
cbx33fungi, is the OpenStack ID, the same as the id?16:38
*** samueldmq has joined #openstack-infra16:38
*** ramashri has joined #openstack-infra16:39
*** afazekas has joined #openstack-infra16:39
sdaguedid we lose a cloud - ?16:40
*** flepied1 has quit IRC16:40
sdaguethe node count went way down, but there is still a lot of stuff in queue16:40
anteayasc68cal: reminder qa meeting in 22 minutes16:40
kragnizcould one of the gerrit admins add me to the initial searchlight groups?16:41
sc68calanteaya: thx16:41
anteayasdague: doesn't look like it is host related16:41
anteayasdague: not sure, did something zuul, nodepool related just merge?16:42
fungicbx33: is the openid provider we maintain for the site (and also and will be eventually moving our other sites to. right now and the rest are using as their openid provider still16:42
* anteaya checks too16:42
*** nfedotov has quit IRC16:42
sdagueanteaya: it could be, the drop in node count was surprising to me for this time of day16:42
openstackgerritAdam Gandelman proposed openstack-infra/project-config: Adds akanda-appliance-buildimage post job
fungisdague: it looks like nodepool stopped building new nodes... i'll see if there's any obvious problem16:43
*** romcheg has joined #openstack-infra16:43
anteayasdague: yes it is16:43
anteayazuul 41 hours ago, nodepool 15 hours ago for latest patches16:43
*** nibalizer has joined #openstack-infra16:43
cbx33fungi just creating my Foundation member16:43
cbx33just wondering if I needed to create an "account" there too16:43
cbx33or if i used my launchpad creds16:43
fungicbx33: it's separate from launchpad. if you've not created an account there you'll need to16:44
cbx33cool ty16:44
*** jasondotstar is now known as jasondotstar|afk16:44
fungicbx33: fairly soon (we hope) all our systems will just need an account on and we'll no longer need people to have accounts on to contribute16:45
greghaynesjeblair: I'm seeing a lot of transient failures in DIB when we try to download various images - if we wanted to make an image hosting mirror do you have any thoughts on where that should live?16:45
*** xyang0 has quit IRC16:45
*** jasondotstar|afk is now known as jasondotstar16:46
*** afazekas has quit IRC16:46
*** jasondotstar is now known as jasondotstar|afk16:46
*** spzala has joined #openstack-infra16:46
fungisdague: nodepoold is logging lots of timeout errors for gearman admin requests to zuul, so presumably isn't aware of the zuul demand and is just satisfying min-ready counts now16:47
cbx33fungi, well...I added my openstack id...they both use the same email addres and it's still not working16:47
*** derekh has quit IRC16:48
fungicbx33: you joined the foundation as an individual member, per the instructions?16:48
cbx33I joined as a community member16:48
*** afazekas has joined #openstack-infra16:49
mordredfungi, sc68cal: I feel like I'm not smart enough with bash to understand something in 19290716:49
mordredwhat do the embedded $ do in the line?16:49
fungicbx33: yep, that won't work. you need to join the foundation as an individual member. i've got a clarification proposed to the instructions to make that more obvious
sc68calmordred: oh I have no idea, I just saw people doing that in other instances where they had multiple lines. Maybe it's a string concat operator?16:50
*** RichardRaseley has joined #openstack-infra16:51
greghaynesI think that will put a literal $ in there?16:51
fungicbx33: see also for a screenshot where you can see the button you need to click to change that16:51
mordredI agree - it looks like it's saying "please concat these two strings using '\n' as the concatenation character16:52
greghaynesoh, im wrong16:52
*** otter768 has joined #openstack-infra16:52
*** puranamr has joined #openstack-infra16:52
mordredmordred@camelot:~/src/openstack-infra/shade$ echo "foo"$'\n'"bar"16:52
mordredneat. I have learned a new thing today16:52
*** jasondotstar|afk is now known as jasondotstar16:53
* cbx33 does the dance16:53
*** jasondotstar is now known as jasondotstar|afk16:53
cbx33thanks fungi16:53
fungisdague: jeblair: looks like the gearman timeouts might coincide with glance image uploads
fungicbx33: you're welcome!16:53
mordredfungi: oh - that's fun16:53
fungicbx33: and sorry that process is so confusing, we're trying to make it simpler, but there are lawyers involved in the changes we need16:54
mordredSpamapS: I verify that your fix to shade fixes the metadata16:54
*** ashleighfarnham has joined #openstack-infra16:55
*** yamamoto has quit IRC16:55
anteayasc68cal: so what that your intent with the $ in the script?16:55
anteayawas that16:55
*** HeOS_ has quit IRC16:56
mordredShrews: if you get a sec, could use +As on 193194 and 193164 so we can release and roll in to prod for nodepool16:56
sc68calanteaya: I'm not very skilled at bash so I just looked what other people were doing to make multiple lines for that variable16:56
*** crc32 has joined #openstack-infra16:56
mordredSpamapS: you too on 19316416:56
mordredsc68cal: +A btw16:56
*** dizquierdo has quit IRC16:56
greghaynesmordred: oh, $'\n' is interpreted as a newline, so its a hack to newline without multiple lines16:56
anteayasc68cal: guess we'll find out16:56
cbx33fungi, w00t
anteayain the experimental logs soon16:57
fungicbx33: congrats and welcome!16:57
anteayamordred: thank you16:57
sc68calmordred: thanks :)16:57
*** otter768 has quit IRC16:57
openstackgerritMerged openstack/requirements: Increase testtools version to support unittest2
fungimordred: i have a feeling that 400mbps we see flat-lining during our dib uploads to hpcloud indicates our bandwidth cap for that flavor16:58
kragnizcould someone add me to the groups of a new project?16:58
fungimordred: and there's a good chance we have poor scheduling or lossy rate limiting causing gearman connectivity to flounder while that's going on16:58
fungikragniz: sure, what was the change that added the new project?16:58
*** baoli has quit IRC16:59
eantyshevHello! May I ask you to add these accounts to CI group: vzstorage-ci and virtuozzo6-ci17:00
fungikragniz: done17:00
kragnizfungi: awesome17:00
kragnizfungi: thanks!17:00
*** baoli has joined #openstack-infra17:00
fungikragniz: you're welcome17:01
*** dpaterson has quit IRC17:01
*** [HeOS] has quit IRC17:01
fungianteaya: gothicmindfood: krtaylor: mmedvede: asselin: if one of you is around, see eantyshev's request above17:02
*** yamamoto has joined #openstack-infra17:02
* asselin looks17:02
jeblairgreghaynes: i think we need one of us to circle around and spiff up afs again; i think that's the best way to have an image mirror17:02
anteayaasselin: thanks, thanks fungi17:02
asselineantyshev, I think I already added you:,members17:03
mmedvedeasselin is fast :)17:03
fungisdague: looks like nodepool managed to get a gearman word in edgewise and noticed zuul has some unmet demand. this will probably continue to be spiky until the image uploads wrap up17:03
sdaguefungi: gotcha17:03
kragnizfungi: could you add me to searchlight-core, also?17:03
fungioh, looks like maybe they just did wrap up17:03
fungikragniz: i thought i did... double-checking now17:03
*** jasondotstar|afk has quit IRC17:03
asselineantyshev, I tend to add ci accounts when I get their unwanted e-mails :)17:04
pabelangerso, this is odd. Stackalytics is cloning the docker repo and trying to import that into its database of statistics17:04
fungisdague: jeblair: pretty sure the start and end of the gearman timeouts for nodepool correlate precisely to the upload start/end bandwidth graph17:04
fungiso that seems to explain it17:04
Clintpabelanger: frightening17:04
eantyshevasselin: thanks, virtuozzo6-ci is there ! please, add vzstorage-ci too17:05
asselineantyshev, but you now have 3 accounts listed there. Let me know if any should be removed:,members17:05
Clintjeblair: oh, btw, my urwid probblem was due to UXTerm*metaSendsEscape: false17:05
eantyshevasselin: one of them (the one with email is deprecated)17:05
asselineantyshev, ok will remove that one. thanks.17:05
fungikragniz: done, sorry about that. i must have pasted and forgotten to hit the add button the first time17:06
kragnizfungi: that's okay! Thanks!17:06
eantyshevasselin: Thank you!17:06
fungiClint: oh, yep, that option is teh evilz17:06
anteayaeantyshev: should we disable the gerrit account with
nibalizerpabelanger: niiiice17:07
mordredpabelanger: it does ansible too17:07
mordredand something else I can't remember17:07
*** ddieterly has joined #openstack-infra17:07
eantyshevanteaya: yes, sure17:07
Clintfungi: i can't imagine why someone flipped it on by default17:07
mordredfungi: perhaps we shoudl figure out how to ratelimit the upload a little better so that it doesn't hog the bandwidth17:07
fungiClint: i've been cargo-culting xterm*metaSendsEscape:true in my .Xresources for so long i'd forgotten what it was for17:07
pabelangermordred, ya, processor is dying on utf8 issues in docker. Guess it is time to setup a filter on cloning repos17:08
jeblairfungi, mordred: oh, wow.  so we're either starving network traffic on nodepool, or the upload process is starving cpu for other threads.17:08
*** rguillebert has quit IRC17:08
openstackgerritMerged openstack-infra/project-config: Fix Manila jobs for running on stable/juno branch
*** achanda has quit IRC17:08
Clintfungi: rxvt-unicode gets it right17:08
* krtaylor reads scrollback17:09
jeblairClint: wow, yeah that looks like "behaveUnlikeAnythingNormal: true"17:09
fungiClint: i think i ended up setting it when i upgraded and discovered that alt+stuff stopped working in weechat17:09
anteayaeantyshev: are you sure that is the email address?17:09
*** kushal has quit IRC17:09
*** kdas_ has joined #openstack-infra17:09
*** yamahata has quit IRC17:09
jeblairmordred, fungi: also, the distributed build/upload workers would help with this17:10
openstackgerritMerged openstack-infra/project-config: Add openstack/designate-dashboard to the DNS Services project
Clintfungi: ah, i always use urxvt for irc17:10
*** kdas_ has quit IRC17:10
anteayaeantyshev: found it17:10
eantyshevanteaya: sorry, pcs-ci-maintainers@parallels.com17:10
anteayapaste.o.o is slow for me to day17:10
Clintof course i also type ESC instead of alt, so..17:10
jeblairmordred, fungi: though no one has signed up to write that yet17:10
*** kushal has joined #openstack-infra17:10
fungiClint: oh, i'm just stockholm-syndromed into sticking with classic xterm for everything (except on my htpc where i use cool-retro-term because it's some serious nostalgic eye candy)17:11
jeblairanyone looking for some python programming? :)17:11
*** achanda has joined #openstack-infra17:11
greghaynesso that would dictate that the distributed builders would have to do the upload themselves17:11
jeblair should be fun hacking17:11
Clintfungi: i tried using urxvt for all terminals at some point and i forget why it didn't take17:11
jeblairgreghaynes: yep; that's the way it's specced at least17:12
greghaynesah, awesme17:12
anteayacan someone please disable gerrit account id 15741:
anteayaeantyshev: I won't send out an email for that17:12
*** bswartz has joined #openstack-infra17:13
fungianteaya: done17:13
anteayafungi: thank you17:13
eantyshevanteaya: It's ok since I'm in charge for that account17:13
anteayaeantyshev: wonderful, it is disabled17:13
*** krtaylor has quit IRC17:14
*** krtaylor has joined #openstack-infra17:14
*** ihrachyshka has quit IRC17:17
*** cdent has quit IRC17:17
*** amotoki_ has quit IRC17:18
*** ryanpetrello has joined #openstack-infra17:18
mordredjeblair, fungi: if you get a sec, could use +As on 193194 and 193164 so we can release and roll in to prod for nodepool17:18
*** ddieterly has quit IRC17:18
fungimordred: we also need to clean up jeblair's alien cleanup patch which is currently hand-applied on nodepool.o.o (where puppet agent is still disabled)17:19
*** ZZelle_ has joined #openstack-infra17:19
fungii'll see where that's sitting17:19
jeblairi thought i cleaned the patch yesterday17:19
*** achanda has quit IRC17:19
jeblairyeah, passes tests now17:19
*** jasondotstar has joined #openstack-infra17:20
*** jasondotstar is now known as jasondotstar|afk17:20
*** jasondotstar|afk is now known as jasondotstar17:20
fungijeblair: awesome. rereviewing so we can release puppet on nodepool17:21
mordreddoing the same17:21
*** cdent has joined #openstack-infra17:21
openstackgerritBenjamin Pflanz proposed openstack-infra/git-review: Don't parse git log header for topic
*** abitha has joined #openstack-infra17:23
yolandamordred, jeblair, about the comment for nodepool double queues and rate.. makes sense if i work on a way to have a shared rate between all task workers? will you be willing to merge that if rate is respected?17:23
*** achanda has joined #openstack-infra17:24
mordredyolanda: well... it gets really screwy with swift image uploads17:24
*** cdent has quit IRC17:24
yolandamordred, what do you mean?17:24
mordredyolanda: because swiftservice handles multi-threaded chunking behind the scenes17:25
openstackgerritAndreas Jaeger proposed openstack-infra/irc-meetings: Revert "Fix incorrect docs meeting time"
mordredyolanda: the swift image upload does a bunch of stuff that does not get accounted for by TaskManager17:25
yolandamm, but it will behave the same for one queue and for two17:25
mordredthat is true17:25
mordredjeblair, fungi: +2 on both17:26
yolandajeblair's concern for that, and he is right, is that my change can be doubling rate17:26
*** jasondotstar is now known as jasondotstar|afk17:26
yolandabut if i have some  way to calculate that rate for x task workers , that problem will be gone17:26
yolandaso if you want to adopt that 2 queues effort i will be more than happy to work on it17:27
mordredyolanda: I need to go get a new bag of coffee - lemme ponder while I do that17:28
fungiyolanda: does my answer to your comment on 190827 make sense?17:29
fungiyolanda: it's of course possible i'm misinterpreting it17:29
*** jasondotstar|afk is now known as jasondotstar17:29
yolandai'm just going out to the cinema, bbl in a pair or hours17:29
*** dguerri is now known as dguerri`17:30
pelixzaro: seems to be doing the right thing with the tests it has, since it exercises using conf files, was there something else that needed to be checked with it?17:32
*** sdake_ has joined #openstack-infra17:32
SpamapSmordred: lol, you guys +A'd a syntax error. ;)17:33
fungimordred: once 190827 merges i'll reenable puppet agent on the server17:33
* SpamapS is just now getting to testing that on his devstack node17:33
fungiSpamapS: what change?17:34
SpamapSit failed verification, I'm fixing it17:34
AJaegerSpamapS: We're human ;(17:34
fungiSpamapS: hah. mordred and Shrews were just eager, that's all ;)17:35
jeblairthat's why we have zuul :)17:35
mordredall hail zuul17:35
openstackgerritClint 'SpamapS' Byrum proposed openstack-infra/shade: Fix MD5 headers regression
*** ramashri_ has joined #openstack-infra17:35
AJaegeryeah, missing closing ")". A good interpreter should add those automatically ;)17:35
ShrewsSpamapS: fungi: we have to leave Jenkins something to do17:35
pleia2AJaeger: would it be possible for you to sync up with StevenK at the end of his day/beginning of your day tomorrow to land 181113?17:36
*** sdake has quit IRC17:36
fungiyeah, this is python. who needs () can't we just add some more indentation somewhere instead?17:36
*** spzala has quit IRC17:36
SpamapSShrews: you're the job creators. #merikuh17:36
AJaegerpleia2: the patch looks fine to go in.17:36
AJaegerpleia2: I'm happy to remove the WIP - I was waiting for your "Zanata is ready for this" ;)17:37
*** sarob has joined #openstack-infra17:37
*** asrangne__ has quit IRC17:37
AJaegerpleia2: shall we exercise it? Then I'll remove my WIP and can +A - or you can +A after talking with StevenK17:37
pleia2AJaeger: ok, you can remove it and I'll confirm with him when he comes on later so we can land it :)17:37
pleia2AJaeger: it was kind of in a perpetual-but-not-marked WIP so I wanted to be sure17:37
SpamapSstupid Ubuntu server installer setting non-VGA modes17:37
*** ramashri has quit IRC17:38
bknudsonI'm trying to figure out why a job is failing sometimes... fails on the new gate-python-keystoneclient-tox-bandit17:38
bknudsonand the error is ERROR: unknown environment 'bandit'17:38
AJaegerpleia2: Since cinerama wrote the patch, StevenK was not able to WIP himself - therefore I did it for him...17:38
bknudsonit does tox -e bandit17:38
bknudsonbut that env does exist since it was merged recently --
mordredSpamapS: you know, you could skip the installer and just used bifrost17:38
openstackgerritMerged openstack-infra/nodepool: Isolate and persist fake client state
bknudsonso maybe the job isn't merging to master?17:39
* mordred hides17:39
pleia2AJaeger: ah right, gotcha :)17:39
*** sarob has quit IRC17:39
SpamapSmordred: I could. I might.17:40
jeblairbknudson: that change is on a feature branch17:40
bknudsonoh, duh17:40
jeblairbknudson: so it's testing merging it into the feature branch17:40
fungibknudson: is that env in the tox.ini on the feature/keystoneauth_integration branch?17:40
AJaegerbknudson: the patch is on a feature branch - does that feature branch also have the bandit environment?17:40
fungier, yeah, what jeblair just said17:40
jeblairand AJaeger17:41
bknudsonI'm not going to worry about it then17:41
AJaegererr, what fungi and jeblair said ;)17:41
jeblairbknudson: we all agree! :)17:41
fungiresounding agreement! our work here is done. let's all go out for drinks17:41
fungibknudson: it's possible openstack-infra/project-config:zuul/layout.yaml needs a (stricter?) branch filter for that job so that it only runs on relevant branches where the tox.ini has the necessary env defined17:42
jeblairbknudson: anyway, you could merge master into the feature branch, or disable that job on the feature branch, or ignore it, or probably some other things i'm thinking of.  not sure if cherry-picking the change across branches would cause merge conflicts later.  hopefully not, but no guarantees.17:43
openstackgerritMerged openstack-infra/nodepool: Delete leaked instances
*** mmedvede has quit IRC17:43
AJaegersince the job is non-voting, it's fine to wait for a merge as well - unless you want to make it voting...17:43
bknudsonI'd expect there'd be a maitainer of the branch that merges with master occasionally.17:44
fungiokay, that's merged, so once the current ansible run is complete, i'm reenabling puppet agent on nodepool.o.o again17:44
*** erlon has joined #openstack-infra17:44
*** yamamoto has quit IRC17:44
fungibknudson: looks like that feature branch was created without updating the keystoneclient acl to enable anyone to push a merge commit17:44
fungibknudson: so right now the set of people doing that is approximately 017:45
*** jeax has joined #openstack-infra17:45
SpamapSlovely... running shade functional tests on my devstack node seems to have killed networking on my devstack node17:45
*** whitenoise has quit IRC17:45
SpamapSor it just went down independently of that17:45
*** jeax is now known as whitenoise17:45
*** sarob has joined #openstack-infra17:46
erlondhellmann: hi Doug, I notice a '7.0.0a0' tag in Cinder branch, what's that about?17:46
openstackgerritMerged openstack-infra/project-config: Run oslo.db unit tests with MySQL-Python too
bknudsonfungi: he he... I guess they really didn't want to keep it up to date.17:46
fungierlon: that's in preparation for the next cinder release not being year.serial based17:47
*** mmedvede has joined #openstack-infra17:47
fungierlon: there's a mailing list thread... let me get it for you17:47
*** sarob has quit IRC17:47
erlonfungi: hmm, missed that, have just returned from vocations, and lost the discution17:48
*** dannywil_ has quit IRC17:50
fungierlon: it starts back in may, but the first message in the june archive for it is
*** openstackgerrit has quit IRC17:50
fungierlon: it's the "[all][release] summit session summary: Release Versioning for Server Applications" thread17:51
*** dannywilson has joined #openstack-infra17:51
*** openstackgerrit has joined #openstack-infra17:51
*** jasondot_ has joined #openstack-infra17:51
dhellmannfungi: can you add me to,members please?17:52
fungierlon: the first message about the alpha tags is
*** marzif has joined #openstack-infra17:52
anteayamordred: if you feel like supporting sc68cal's devstack documentation efforts:
erlonfungi: hmmm, grea, thanks, Ill see that17:52
fungidhellmann: done17:52
openstackgerritMerged openstack-infra/shade: Ensure that service values are strings
*** jasondot_ has quit IRC17:53
dhellmannerlon: fungi : thanks17:55
openstackgerritwill soula proposed openstack-infra/jenkins-job-builder: Add support for the Naginator Plugin
*** tvanderwiel has quit IRC17:56
*** SumitNaiksatam has joined #openstack-infra17:56
marzifanyone could help moving forward:
openstackgerritSpencer Krum proposed openstack-infra/puppet-logrotate: Make logrotate::file more resilient
*** signed8bit has quit IRC17:59
*** vinayyadhav has quit IRC17:59
*** vinay_yadhav has quit IRC17:59
*** ivar-lazzaro has joined #openstack-infra18:00
*** ivar-lazzaro has quit IRC18:00
*** rkukura has joined #openstack-infra18:01
*** ivar-lazzaro has joined #openstack-infra18:01
*** ramashri_ has quit IRC18:03
*** ramashri has joined #openstack-infra18:03
*** ildikov_ has quit IRC18:04
*** spzala has joined #openstack-infra18:04
*** tosky has quit IRC18:04
*** tonytan4ever has quit IRC18:06
*** jasondotstar is now known as jasondotstar|afk18:07
*** hodos has quit IRC18:07
*** tonytan4ever has joined #openstack-infra18:07
*** afazekas has quit IRC18:08
pelixany infra core reviewers fancy reviewing for JJB?18:11
*** tonytan4ever has quit IRC18:11
*** sarob has joined #openstack-infra18:12
*** baoli has quit IRC18:13
*** sarob has quit IRC18:13
*** samueldmq2 has joined #openstack-infra18:14
*** cdent has joined #openstack-infra18:14
*** samueldmq__ has joined #openstack-infra18:15
*** achanda has quit IRC18:15
*** jasondotstar|afk is now known as jasondotstar18:15
*** jasondotstar is now known as jasondotstar|afk18:15
AJaegerI don't think the designate-dashboard ACLs are setup correctly ;(18:17
*** samueldmq has quit IRC18:17
*** sarob has joined #openstack-infra18:17
AJaegerLooking at gerrit/acls/openstack/designate-dashboard.config I see "inheritFrom = openstack/designate" but looking at,access I see only horizon-core.18:18
AJaegerfungi, jeblair, could you check this, please? ^18:18
*** samueldmq2 has quit IRC18:18
*** tonytan4ever has joined #openstack-infra18:18
*** amotoki_ has joined #openstack-infra18:18
fungiAJaeger: see above that where it says "rights inherit from"18:20
*** samueldmq2 has joined #openstack-infra18:22
*** sarob has quit IRC18:22
*** pelix has quit IRC18:23
*** amotoki_ has quit IRC18:23
*** samueldmq has joined #openstack-infra18:23
jeblairer should we be using inheritfrom?18:24
AJaegerfungi: ah, missed that line... Thanks!18:24
fungiif i had reviewed that change i would have recommended just being explicit in the acl18:24
jeblairi vote we make it standard18:25
AJaegerfungi: I reviewed, sorry ;(18:25
*** signed8bit has joined #openstack-infra18:25
AJaegerjeblair: shall I send a cleanup patch?18:25
*** samueldmq__ has quit IRC18:25
jeblairAJaeger: that would be great.  that way it's easier to read and understand, and our tools don't have to be as complex18:25
fungiinherit from makes the acl harder to grok, and doesn't really add much convenience when teh acls are under config management anyway18:26
*** samueldmq2 has quit IRC18:26
AJaegerthanks for the explanation, will not let it through again18:27
fungiAJaeger: might not hurt to grep for any other similar acls which have slipped through and fix those too18:27
fungiit's possible that was cargo-culted from another acl file as acceptable18:27
AJaegerfungi: will do...18:27
*** jasondotstar|afk has quit IRC18:27
fungialso, nodepoold is back in sync with master again18:28
*** jtomasek has quit IRC18:28
fungiand puppet is working without error18:28
AJaegerfungi: no cargo-cult, this is the only one...18:28
fungiAJaeger: excellent--thanks!18:28
*** samueldmq2 has joined #openstack-infra18:30
*** xyang0 has joined #openstack-infra18:31
*** ramashri has quit IRC18:31
openstackgerritChristopher Aedo proposed openstack-infra/system-config: Add openstack-app-catalog mailing list
*** kdas_ has joined #openstack-infra18:33
openstackgerritMerged openstack-infra/project-config: Run python34 tests only on master for heat
*** dboik has quit IRC18:34
*** dboik has joined #openstack-infra18:34
*** samueldmq has quit IRC18:34
*** koolhead17 has joined #openstack-infra18:34
*** kushal has quit IRC18:35
*** ildikov_ has joined #openstack-infra18:36
*** claudiub has quit IRC18:36
*** ramashri has joined #openstack-infra18:37
*** marzif has quit IRC18:37
*** hdd has quit IRC18:37
*** samueldmq__ has joined #openstack-infra18:38
*** thedodd has quit IRC18:38
*** ajmiller has quit IRC18:38
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Remove inherit-from from designate-dashboard ACL
AJaegerfungi, jeblair: here's the promised patch ^18:39
*** pblaho has quit IRC18:39
fungiAJaeger: thanks again18:39
*** samueldmq has joined #openstack-infra18:39
*** marzif has joined #openstack-infra18:40
*** samueldmq2 has quit IRC18:42
*** samueldmq__ has quit IRC18:42
*** [HeOS] has joined #openstack-infra18:42
*** samueldmq2 has joined #openstack-infra18:44
*** hdd has joined #openstack-infra18:44
*** samueldmq__ has joined #openstack-infra18:44
*** yamamoto has joined #openstack-infra18:45
*** marzif has quit IRC18:47
*** samueldmq has quit IRC18:47
*** notnownikki has quit IRC18:48
*** jtomasek has joined #openstack-infra18:48
cdentI'm seeing this in a few different places:
*** samueldmq2 has quit IRC18:48
lifelessfungi: thanks :)18:48
cdentwhat does it tell me?18:48
*** samueldmq has joined #openstack-infra18:48
lifelessSpamapS: for amazing, look at the HTTP/1.1 definition of header :)18:49
*** Sukhdev has joined #openstack-infra18:49
cdent(not just gnocchi)18:49
*** [HeOS] is now known as HeOS18:49
lifelesssdague: we could put it in a venv easily enough and call it via the venv's python18:50
lifelesssdague: would you like me to whip that up ?18:50
*** yamamoto has quit IRC18:50
lifelesssdague: it is itself consistent with global-requirements, so it can't interfere withthe pinning process18:50
openstackgerritMerged openstack-infra/project-config: Add a non-voting py34 check job to Glance
lifelessSpamapS: also you're welcome (wheel caching :))18:51
*** samueldmq__ has quit IRC18:51
*** samueldmq2 has joined #openstack-infra18:51
openstackgerritChristopher Aedo proposed openstack-infra/system-config: Add app-catalog-admin mailing list
*** samueldmq__ has joined #openstack-infra18:54
mordredSpamapS: swiftclient.service.SwiftError: "Metadata parameter x-object-meta-x-shade-sha256 must contain a ':'.\nExample: 'Color:Blue' or 'Size:Large'"18:54
*** samueldmq has quit IRC18:55
*** cdent has quit IRC18:55
jeblairbut no space?18:55
*** samueldmq has joined #openstack-infra18:55
*** dboik has quit IRC18:55
mordrednotmyname: that's an error we just got in our functional tests of our swift image uploads18:56
notmynamewhat's the command?18:56
notmynameor call?18:56
mordrednotmyname: one sec ...18:56
*** dimtruck is now known as zz_dimtruck18:56
*** samueldmq2 has quit IRC18:57
*** zz_dimtruck is now known as dimtruck18:57
*** samueldmq2 has joined #openstack-infra18:57
*** samueldmq__ has quit IRC18:58
*** mwagner_afk has quit IRC18:58
*** ociuhandu has quit IRC18:59
notmynamemordred: looks like the "header" list in the upload() call must be a list of 'Key: Value" strings18:59
notmynamemordred: eg ['x-shade-sha256: deadbeef']18:59
mordrednotmyname: it's doing swift_service.upload(container=container, objects=[SwiftUploadObject(source=filename, object_name=name)], options=dict(headers= ...18:59
mordredok - we were passing in a dict18:59
notmynamemordred: I'm still looking. let me see what else I suss out19:00
*** samueldmq has quit IRC19:00
*** samueldmq2 is now known as samueldmq19:00
*** _nadya_ has joined #openstack-infra19:00
*** samueldmq2 has joined #openstack-infra19:01
*** jasondotstar has joined #openstack-infra19:01
pabelangerIs it possible to have a guest / anonymous user connect to gerrit via SSH?19:01
notmynamemordred: ah, wait. are you sending extra headers or tryignt o set metadata?19:01
*** jasondot_ has joined #openstack-infra19:01
*** dboik has joined #openstack-infra19:01
*** dannywilson has quit IRC19:01
mordrednotmyname: trying to set metadata19:01
*** dboik has quit IRC19:01
nibalizerpabelanger: I don't think so, I've been using service users which is :(19:02
mordrednotmyname: but doing it by sending headers - so maybe that's not the right interface19:02
*** dboik has joined #openstack-infra19:02
*** jasondotstar has quit IRC19:02
*** claudiub has joined #openstack-infra19:02
*** dannywilson has joined #openstack-infra19:02
mordrednotmyname: the header addition worked against rackspace - so maybe swift master has gotten stricter?19:02
pabelangernibalizer, well, there is a RESTapi since gerrit 2.5 that gives what I need (projects). So, I'll likely use that for my needs19:04
*** kdas_ has quit IRC19:04
mordrednotmyname: but - ultimately, we want to add 2 pieces of metadata to the object, and would prefer to do it at upload time, rather than as a follow-on POST19:04
openstackgerritMerged openstack-infra/project-config: Enable horizon integration tests voting
pabelangermostly save me the hassel of adding a temp SSH key to my account19:04
*** samueldmq has quit IRC19:05
*** _nadya_ has quit IRC19:05
*** samueldmq2 has quit IRC19:05
nibalizercan you RESTapi the stream-events somehow?19:06
pabelangernot sure19:06
pabelangeryou'd need some WS for that19:06
pabelangeror long poll HTTP19:06
*** jasondot_ is now known as jasondotstar|afk19:07
*** hashar has joined #openstack-infra19:07
asselinI'm looking for a way to abort/dequeue a job when a patch merges. I didn't find anything configuration to do so. I submitted this WIP patch, but would like to know if there's a way to do it w/o it:
*** linkedinyou has joined #openstack-infra19:09
notmynamemordred: yeah, that's the right way (set meta on upload rather than after). where right==more efficient19:09
*** ramashri has quit IRC19:09
mordrednotmyname: awesome. so - for this, I really want to be passing to meta rather than headeres, yeah?19:10
notmynamemordred: also, wow, this is *such* an easy codebase to read! /s19:10
rcarrillocruzpabelanger: have you tried with the internal 'Gerrit Code Review' user19:10
*** ramashri has joined #openstack-infra19:10
notmynamemordred: well, that's what I was looking for. I can see how it works with a post witht eh meta[] but not on upload[]19:10
pabelangerrcarrillocruz, which one is that?19:11
mordrednotmyname:                 headers = split_headers(19:11
notmynamemordred: so for now, I'd say use header[] but include the x-object-meta- prefix and it should work19:11
mordred                    options['meta'], 'X-Object-Meta-')19:11
*** romcheg has quit IRC19:11
mordrednotmyname: it looks like meta is just syntactic sugar for headers to not need the prefix perhaps?19:11
rcarrillocruzessentially, it's an internal user that allows you to do a 'su', and do things as other users19:11
mordredbut easiest thing for us would be to do what you jsut said19:11
mordredand just keep the x-object-meta prefix19:11
*** hdd has quit IRC19:11
mordredsince we've got it written that way now19:11
notmynamemordred: right. and that line you found is in post() not upload()19:11
mordredoh - right19:11
notmyname(because that makes sense)19:12
lifelessAJaeger: +A would be fine now IMO19:12
pabelangerrcarrillocruz, pretty sure I still need a private key19:12
lifelessAJaeger: we won't know if its completely right until all the bits are landed, but a failing periodic job is better than guesswork19:12
pabelangerwhich, I don't have access to19:12
rcarrillocruzyes, it's the gerrit2 ssh key19:12
rcarrillocruzi.e. /home/gerrit2/review_site/etc/ssh_host_yadayada19:13
pabelangerya, this would be for stackalytics host19:13
notmynamemordred: yeah, so header = ['x-object-meta-x-share-sha256:deadbeef'] would work for you, I think19:13
lifelessmordred: is urgent.19:13
pabelangerso, it could be setup to use it19:13
notmyname*shade, not share19:13
pabelangerbut doesn't help right now, since I would need to get access to it to run locally19:13
pabelangerbut thanks19:13
*** wenlock has joined #openstack-infra19:14
pabelangerHonestly, RESTapi works better for me19:14
mordrednotmyname: AWESOME19:14
pabelangerI should add support for it19:14
lifelessfungi: dhellmann: mordred: and I'd *very much* like to get the stack up to in19:14
*** ddieterly has joined #openstack-infra19:14
*** ddieterly has quit IRC19:14
*** dangers is now known as dangers_away19:16
fungilifeless: ahh, yep, just got back to reviewing more of that stack now that i've got my pymysql changes going for a while19:16
*** sdake has joined #openstack-infra19:17
jeblairpabelanger: yeah, ++ rest api as long as you don't need stream-events (and i don't think stackalytics does19:18
*** romcheg has joined #openstack-infra19:18
*** sarob has joined #openstack-infra19:18
fungidougwig: what process did you follow with ? i see signs that fwaas may be in a similar situation19:19
*** otter768 has joined #openstack-infra19:19
pabelangerjeblair, first step would be to port stackalytics to gerritlib, then add REST support it gerritlib19:19
pabelangerseems like a no brainer19:19
*** dustins_ has joined #openstack-infra19:20
sdaguelifeless: yeh, installing into a dedicated venv seems safer19:21
*** hdd has joined #openstack-infra19:21
*** sdake_ has quit IRC19:21
sdaguejust so it doesn't accidentally impact the rest of the system19:21
jeblairpabelanger: could do; but if you're doing unauthenticated access, it's only a couple of lines with requests19:21
lifelesssdague: Its not clearto me how it can19:21
lifelesssdague: its deps are also in global-requiremetns19:21
lifelesssdague: and we should put it in openstack/requirements too19:22
sdaguelifeless: so here is the failure scenario I can imagine19:22
lifelesssdague: regardless of venv or not.19:22
sdagueit ends up requiring something19:22
sdaguea project ends up using that thing19:22
*** dustins has quit IRC19:22
pabelangerjeblair, right. Was being ambitious with my requirements19:22
sdaguebut doesn't have it in it's requirements file, because it works without it19:22
fungijeblair: pabelanger: unfortunately if it wants access to methods you wouldn't think should need auth like the /groups tree then it gets more complex19:22
sdaguein our integration environment19:22
sdaguebut in a packaged environment, there are no such guaruntees19:22
jeblairfungi: oh interesting19:22
lifelesssdague: but that failure exists for everything today19:23
pabelangerfungi, thanks for the heads up.  Need to check the code to see if it does19:23
fungijeblair: yeah, they've made some interesting decisions about what requires authentication, mostly because they want to do rbac to determine who gets to see what information19:23
sdaguelifeless: it does, but this adds one more surface for that19:23
jeblairpabelanger: if you find that you need auth access, then i would recommend updating gerritlib.  pull code from gertty into it; it's has pretty robust auth handling at this point.19:23
sdagueand a pretty non obvious one19:23
*** tvanderwiel has joined #openstack-infra19:23
*** sarob has quit IRC19:23
pabelangerjeblair, Oh, nice. I wanted to know how gertty did it, and planned to look19:24
*** otter768 has quit IRC19:24
sdagueanyway, we can sort it out at the end, but this is a thing we should consider19:24
lifelesssdague: tis your call. I can whip upa  venv thing19:24
lifelesssdague: Personally I don't rate the incremental risk significant, but - I'll whip up the patch and you can review19:24
jeblairpabelanger: (gertty started as "a couple of lines around requests" but as people use it with more varied gerrits, auth handling has gotten more complex -- enough so that librarification is a good option now.  of course stackalytics only needs to work with one, but who knows how we will change it in the future)19:25
sdagueI think a venv patch would be nice, I'm also happy considering that a tidying up once the bulk of the rest of it is done19:25
*** thedodd has joined #openstack-infra19:25
lifelesssdague: I'll whip it up while waiting for reviews :)19:25
pabelangerjeblair, okay. I'll look into it and see what is needed19:25
sdaguelifeless: :)19:26
*** mrunge has quit IRC19:26
*** jasondotstar|afk is now known as jasondot_19:27
sdaguelifeless: I will attempt to do some reviews19:27
openstackgerritMonty Taylor proposed openstack-infra/shade: Fix MD5 headers regression
*** jasondot_ is now known as jasondotstar|afk19:27
*** jasondotstar|afk has quit IRC19:27
mordredSpamapS, notmyname, jeblair, Shrews, fungi: ^^ check me on that19:27
mordredlifeless: looking19:28
mordredlifeless: oh good - it's already done19:28
jeblairmordred: you're sure it should be "key:value" not "key: value" ?19:28
mordredjeblair: yes19:29
jeblairmordred: the great thing about standards is we can just ignore them!19:29
mordredjeblair: according to both the error message and notmyname19:29
notmynameuh I don't think it matters, but I'll check19:29
mordredjeblair: swiftclient.service.SwiftError: "Metadata parameter x-object-meta-x-shade-sha256 must contain a ':'.\nExample: 'Color:Blue' or 'Size:Large'"19:30
jeblairmordred: yeah, i remember that.  i'm still balking at it.19:30
notmynamehmm.. I don't see a strip(' ') where I'd expect it to be, so best leave it out for now19:30
*** sarob has joined #openstack-infra19:31
jeblairnotmyname: it'd be nice to have the option of using an rfc compliant header :)19:31
mordredor a dictionary...19:32
fungiyou might remember headers from such popular ietf publications as "rfc 822," "rfc 2822" and "jon postel's greatest hits"19:32
*** sdake_ has joined #openstack-infra19:32
openstackgerritMatt Riedemann proposed openstack-infra/elastic-recheck: Update query for cinder notifications race bug 1412513
openstackbug 1412513 in Cinder "Intermittent unit test failures for incorrect notifier count" [High,Confirmed] - Assigned to Eric Harney (eharney)19:33
notmynamejeblair: yeah, it seems weird to me too. however, this is to generate the header value for a reuqest to swift. and swift doesn't store the surrounding whitespace19:33
jeblairfungi: but wait! there's more!  we'll throw in rfc1036, rfc1123, rfc 1327, and rfc1496 if you order now!19:34
jeblairnotmyname: so headers in headers?19:34
notmynamejeblair: I don't follow19:34
jeblairi guess i don't either19:35
*** _nadya_ has joined #openstack-infra19:35
notmynamejeblair: WE'RE IN TOTAL AGREEMENT!!!19:35
notmynamemordred: I would totally support a bug report on that sort of stuff19:35
fungiso sounds like maybe swift would handle it sanely but swiftclient is making silly assumptions19:36
*** jasondotstar has joined #openstack-infra19:36
*** sdake has quit IRC19:36
mordrednotmyname: cool. I will do such a thing19:37
*** maishsk_afk has joined #openstack-infra19:37
notmynamefungi: yeah. ultimately, it's the SwiftService thing generating a set of headers which ends up being a dictionary (as one would expect) passed to an http connection19:37
notmynamefungi: so I think it's poor docs and flexibility in swift service19:38
SpamapSWait what?19:39
SpamapSoh never mind19:40
SpamapSI was scrolled back19:40
AJaegerlifeless: ok, let's try it out19:41
greghaynesSpamapS: Welcome to the present19:42
greghaynesSpamapS: Any progress in infra-cloud land?19:42
jeblairmordred: looks like marginally-related test fixes in there.19:43
jeblairmordred: i like them (checking specific shas); but wanted to double check you meant to inclued them19:43
openstackgerritlifeless proposed openstack/requirements: Remove hacking and hacking-owned dependencies.
openstackgerritlifeless proposed openstack/requirements: Mark as Python2.7 only the dependencies that are.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt.
openstackgerritlifeless proposed openstack/requirements: Participate in requirements syncing
mordredjeblair: they're actually required19:43
*** Longgeek_ has quit IRC19:43
KiallAJaeger: re the designate-dash translation change you put up, are you sure the setup.cfg should have designate-dashboard (vs designatedashboard)?
mordredjeblair: because we're stitching strings now19:43
AJaegerKiall: yes!19:43
mordredjeblair: so I have to have a value to put into the string because we need to check the entire string not just the dict key19:44
KiallAJaeger: mind me asking why? the paths don't have a hyphen?19:44
AJaegerKiall: see the commit message - I tested with a dry-run of our infra scripts19:44
AJaegerKiall: Our scripts expect $repo/locale for the file name19:44
lifelesssdague: is there a virtual abstraction in devstack already ?19:44
mordredjeblair: it would have been harder to not just fix the test end-to-end19:44
KiallAh, repo rather than module19:44
*** mnestratov has quit IRC19:44
mordredSpamapS: yah - you likely want to check that change too19:44
sdaguelifeless: ??19:44
jeblairmordred: oh i see that now, on 361.  missed the {/[ change.  even though it was in bold.19:44
mordredSpamapS: oh - you did19:45
mordredjeblair: it's easy to do19:45
*** ramashri has quit IRC19:45
jeblairmordred: i will approve that change as many times as it takes.19:45
AJaegerKiall: exactly19:45
*** dprince has quit IRC19:45
*** ramashri has joined #openstack-infra19:45
KiallHumm.. That's a tad annoying to have both designatedashboard and designate-dashboard, folders, but OK.. Just didn't click that $project was the repo rather than the module.19:46
*** r-daneel has joined #openstack-infra19:47
*** jasondotstar is now known as jasondotstar|afk19:47
AJaegerKiall: that's the way *all* OpenStack projects that translate are setup.19:47
*** sdake has joined #openstack-infra19:47
AJaegerIf you change this, it complicates our scripts because we would need to special case...19:47
SpamapSmordred: testing it right now19:47
AJaegerSo, why not change designatedashboard to designate-dashboard ;)19:48
sdaguelifeless: so we have some venv support, but it assumes it's things that happen after gr is there, because the idea that gr itself would have to be installed was never in the design19:48
KiallAJaeger: yea, it's fine.. Just my OCD will have me eyeballing the extra folder the whole time :P19:49
*** maishsk_afk has quit IRC19:49
SpamapSmordred: works on my devstack19:49
mordredSpamapS: ship it19:49
mordredSpamapS: you know - our functional tests caught something!19:49
*** maishsk_afk has joined #openstack-infra19:49
SpamapSmordred: damn straight19:49
mordredSpamapS: I mean, they didn't catch the other thing19:50
*** baoli has joined #openstack-infra19:50
mordredbut they caught this thing19:50
mordredwhich is exciting19:50
SpamapSmordred: well thats because we weren't asserting enough19:51
*** sdake_ has quit IRC19:51
SpamapSmordred: now we actually are asserting that a file you just uploaded isn't stale. :)19:51
SpamapShm wait19:51
SpamapSfunc tests failed on a second run19:51
*** maishsk_afk has quit IRC19:51
*** jasondotstar|afk is now known as jasondotstar19:52
SpamapShelps if you run with the actual patch checked out19:52
SpamapSmordred: one more fix19:52
openstackgerritClint 'SpamapS' Byrum proposed openstack-infra/shade: Fix MD5 headers regression
*** achanda has joined #openstack-infra19:54
* SpamapS kind of wishes there was a way to have a gate/check job that doesn't have to run the whole of :-P19:55
mordredSpamapS: yah - it would be useful for consumer things19:55
mordredSpamapS: test-prep == "create domain in cloud for test" ; teardown == "disable domain, then delete domain"19:56
mordredsince all cloud objects are contained in a domain19:56
SpamapSprecisely what I was thinking yeah19:56
SpamapSmordred: I wonder if we could do that with infra-cloud19:56
mordredSpamapS: it is a thing I think we'd have the tools to think about once zuul v3 is here19:57
SpamapSlike, here's your temporary domain and creds, your quota is 2 tiny instances. Enjoy.19:57
bswartzzul: ping19:58
*** rfolco has quit IRC19:58
mordredyah - all that's needed is keystone v319:58
*** rfolco has joined #openstack-infra19:59
*** hdd has quit IRC20:00
*** armax has quit IRC20:01
*** baoli has quit IRC20:01
*** baoli has joined #openstack-infra20:02
*** maishsk_afk has joined #openstack-infra20:02
*** wenlock has quit IRC20:02
*** dboik has quit IRC20:03
*** dboik has joined #openstack-infra20:04
*** maishsk_afk has quit IRC20:05
*** jasondotstar is now known as jasondotstar|afk20:06
*** jasondotstar|afk is now known as jasondotstar20:06
jeblairmordred, SpamapS: that does sound cool, but i'm going to be typically cautious about exposing something like that until we figure out how to do it in a robust way that doesn't completely depend on infra-cloud (since the entirety of infra-cloud is one component in an HA system)20:06
*** jasondotstar is now known as jasondotstar|afk20:06
mordredjeblair: yup. totally agree20:06
jeblairmordred, SpamapS: but that will be fun to look at in the future :)20:06
*** markmcclain has quit IRC20:07
mordredjeblair: I think it's a thing we can think about when we have a) zuul v3 and b) more than one cloud that has keystone v320:07
*** ramashri has quit IRC20:07
zulbswartz: hi20:07
*** ramashri has joined #openstack-infra20:07
*** markmcclain has joined #openstack-infra20:08
*** jasondotstar|afk is now known as jasondotstar20:08
openstackgerritMerged openstack-infra/project-config: Add periodic propose-requirements-constraints jobs
*** reed has quit IRC20:09
dougwigfungi: i think i used an fgrep, then just ran tox and CI to see what had imports that didn't match it's package name.  I can do fwaas if needed.20:09
*** tonytan4ever has quit IRC20:11
fungidougwig: it's not urgent, just came to my attention for a similar reason as lbaas (i.e., it declares a dependency on MySQL-python which i don't think it's using because i found no mention of the mysqldb module in the source)20:11
*** tonytan4ever has joined #openstack-infra20:11
*** ramashri has quit IRC20:11
*** reed has joined #openstack-infra20:11
dougwigfungi: i expect the requirements overages happened during the repo split. likely just pulled in neutron's files.20:12
fungidougwig: if it _really_ does need MySQL-python then i'd like to do whatever's necessary to switch it to PyMySQL, but i seriously doubt it does20:12
dougwigfungi: i also doubt it. all of the db stuff is still in neutron.20:12
dougwigfungi: one sec, i can at least get that reference cleanup in progress.20:13
fungidougwig: oh, appreciated! if you confirm it's really not using that particular package i'll abandon 18438420:14
*** ayoung-afk is now known as ayoung20:14
*** maishsk_afk has joined #openstack-infra20:15
*** puranamr has quit IRC20:16
*** tonytan4ever has quit IRC20:16
*** sarob has quit IRC20:16
*** koolhead17 has quit IRC20:19
*** puranamr has joined #openstack-infra20:19
*** abregman has quit IRC20:20
*** sarob has joined #openstack-infra20:20
dougwigfungi: i can't find any references, and i put up , so we'll know in about 45 minutes.  :)20:20
fungidougwig: thanks!20:21
*** kgiusti has left #openstack-infra20:22
dougwigfungi: just did that in vpnaas as well, fyi.20:23
*** puranamr has quit IRC20:24
*** tonytan4ever has joined #openstack-infra20:24
*** puranamr has joined #openstack-infra20:24
davidlenwelljeblair, mordred this should be ready to look at again..
*** jtomasek has quit IRC20:25
*** ildikov_ has quit IRC20:26
fungiSpamapS: you've now angered the pep8 gods with your blasphemous indentation in 19319420:26
*** ldnunes has quit IRC20:26
*** mwagner_afk has joined #openstack-infra20:26
*** hdd has joined #openstack-infra20:26
*** dustins_ has quit IRC20:27
*** baoli has quit IRC20:27
*** puranamr has quit IRC20:28
*** baoli has joined #openstack-infra20:28
*** ldnunes has joined #openstack-infra20:28
SpamapSfungi: haha20:29
*** dustins has joined #openstack-infra20:29
fungiwave a spacebar over it and be absolved20:29
*** rossella_s has quit IRC20:30
*** achanda has quit IRC20:30
SpamapSand that happened because I accidentally borked my vim flake8 plugin last week20:31
*** rossella_s has joined #openstack-infra20:31
SpamapSsilly pathogen20:31
fungiit's a poor craftsman who blames his tools20:31
openstackgerritClint 'SpamapS' Byrum proposed openstack-infra/shade: Fix MD5 headers regression
SpamapSfungi: :|20:31
* fungi is a _very_ poor craftsman, and therefore knows what he's talking about here20:31
* SpamapS has seen fungi's tools .. ;)20:32
sdagueSpamapS: thanks for the devstack patch for mysql, +220:32
*** timrc_ has quit IRC20:32
* fungi was probably drinking heavily20:32
*** abregman has joined #openstack-infra20:33
jeblairdavidlenwell: great, thanks!20:36
*** hdd has quit IRC20:36
*** hdd has joined #openstack-infra20:37
*** sarob has quit IRC20:38
*** ildikov_ has joined #openstack-infra20:39
*** sarob has joined #openstack-infra20:39
*** sarob has quit IRC20:40
*** puranamr has joined #openstack-infra20:41
*** marzif has joined #openstack-infra20:41
*** bcourt has quit IRC20:42
*** sambetts has quit IRC20:43
*** puranamr has quit IRC20:43
*** Longgeek has joined #openstack-infra20:44
*** jog0 has joined #openstack-infra20:44
openstackgerritDoug Hellmann proposed openstack-infra/project-config: Set tag ACLs for more networking libraries
*** maishsk_afk has quit IRC20:45
*** sambetts has joined #openstack-infra20:46
*** tlian has quit IRC20:46
*** jog0 has quit IRC20:47
*** _nadya_ has quit IRC20:48
*** stevemar has quit IRC20:48
*** MarkAtwood has quit IRC20:48
*** fallenpegasus has quit IRC20:48
*** Sukhdev has quit IRC20:49
*** sarob has joined #openstack-infra20:50
*** puranamr has joined #openstack-infra20:51
openstackgerritDoug Hellmann proposed openstack-infra/infra-manual: Update gerrit ACL docs to mention library-release
*** puranamr has quit IRC20:54
*** maishsk_afk has joined #openstack-infra20:54
*** puranamr has joined #openstack-infra20:54
*** fdegir has quit IRC20:54
*** tlian has joined #openstack-infra20:57
marzifany core available to review  we need to split the main repo20:57
tonybjlvillal: Ahhh yeah I suspect the ServerAlias/ServerName is a problem there.20:57
*** puranamr has quit IRC20:58
*** e0ne has joined #openstack-infra20:59
*** wenlock has joined #openstack-infra21:00
*** tlian2 has joined #openstack-infra21:00
openstackgerritKyle Mestery proposed openstack-infra/project-config: Migrate ENABLED_SERVICES and TEMPEST_REGEX into networking-odl
*** tlian has quit IRC21:01
*** jamielennox is now known as jamielennox|away21:01
*** ildikov_ has quit IRC21:02
*** marzif has quit IRC21:03
*** rfolco has quit IRC21:03
*** ildikov_ has joined #openstack-infra21:03
openstackgerritJeremy Stanley proposed openstack-infra/project-config: Correct parameters for generate-constraints
fungilifeless: ^21:04
*** sarob has quit IRC21:05
*** dboik has quit IRC21:06
fungilifeless: also i found what i believe to be a problem in 18924921:06
*** aysyanne has quit IRC21:07
*** tlian2 has quit IRC21:07
fungisee comment21:07
*** jamielennox|away is now known as jamielennox21:09
*** marcusvrn has quit IRC21:11
*** cody-somerville has joined #openstack-infra21:11
*** dboik has joined #openstack-infra21:11
fungiaside from those tweaks, seems to work fine (as far as i can tell without installing the headers needed for every package in global reqs)21:12
*** maishsk_afk has quit IRC21:12
*** maishsk_afk has joined #openstack-infra21:13
fungialso, that reminds me i still need to replace with an instance running trusty21:13
fungii'll try to find time to work on that tomorrow21:13
*** linkedinyou has quit IRC21:13
*** sarob has joined #openstack-infra21:13
lifelessfungi: ack, looking21:14
lifelesscoreycb has encountered a failure in master21:14
lifelesstrying to see if its still a problem or not21:14
*** puranamr has joined #openstack-infra21:14
openstackgerritlifeless proposed openstack/requirements: Remove the shebang from generated's.
openstackgerritlifeless proposed openstack/requirements: Pull serialisation out of _sync_requirements_file.
openstackgerritlifeless proposed openstack/requirements: Remove spurious output in test suite.
openstackgerritlifeless proposed openstack/requirements: Remove hacking and hacking-owned dependencies.
openstackgerritlifeless proposed openstack/requirements: Also update extras in setup.cfg.
openstackgerritlifeless proposed openstack/requirements: Handle : as well as ; for markers.
openstackgerritlifeless proposed openstack/requirements: Participate in requirements syncing
openstackgerritlifeless proposed openstack/requirements: Mark as Python2.7 only the dependencies that are.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt.
openstackgerritlifeless proposed openstack/requirements: Fix setup.cfg.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt generator.
*** ldnunes has quit IRC21:15
lifelessfungi: can you reapply the +A on - was a flaky test failure, but since nothing was in zuul I rebased onto trunk anyhow21:16
*** puranamr has quit IRC21:17
lifelessAJaeger: <_ follow up to the one you acked earlier :)21:17
*** xyang0 has quit IRC21:17
*** baoli has quit IRC21:17
*** baoli has joined #openstack-infra21:18
*** Longgeek has quit IRC21:19
openstackgerritlifeless proposed openstack/requirements: Remove hacking and hacking-owned dependencies.
openstackgerritlifeless proposed openstack/requirements: Participate in requirements syncing
openstackgerritlifeless proposed openstack/requirements: Mark as Python2.7 only the dependencies that are.
lifelessfungi: fixtures added and pushed21:19
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt generator.
openstackgerritMerged openstack-infra/shade: Fix MD5 headers regression
*** otter768 has joined #openstack-infra21:20
fungilifeless: is there a reason to leave that in test-requirements.txt?21:21
*** dkranz has quit IRC21:21
lifelessfungi: meh?21:21
fungiseems like we should avoid duplicating entries in both requirements.txt and test-requirements.txt21:21
fungiunless there's benefit to doing so, i always assumed the test requirements were a superset of the run-time requirements21:22
openstackgerritlifeless proposed openstack/requirements: Remove hacking and hacking-owned dependencies.
openstackgerritlifeless proposed openstack/requirements: Participate in requirements syncing
openstackgerritlifeless proposed openstack/requirements: Mark as Python2.7 only the dependencies that are.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt generator.
mordredWALL OF TEXT21:22
lifelessfungi: its much of a muchness21:22
* mordred goes to review lifeless most recent sequence21:22
lifelessmordred: so, if you review and +A a bunch, it won't be a wall of text anymore :)21:22
lifelessyes, that21:23
fungiyeah, i've already +2'd a few of them, for whatever that's worth21:23
mordredlifeless: I  mean, it will - it'll just be a different wall of text21:23
*** otter768 has quit IRC21:25
*** unicell has quit IRC21:25
*** peristeri has quit IRC21:25
lifelessfungi: its worth progess21:26
fungidhellmann: dims: can you have a look at and weigh in on whether that's going to be a problem conceptually? we can further rule out libs which aren't listed in the projects.txt and those which don't use any of the non-py3k-compliant deps listed in global reqs21:27
*** EricGonczer_ has quit IRC21:27
*** puranamr has joined #openstack-infra21:29
jeblairmordred: thoughts on ansible, nova inventory, launch-node, and onMetal?21:35
*** ddieterly has joined #openstack-infra21:36
jeblairmordred: specifically, if i spin up a couple of onmetal servers, is it going to be a problem with where we are now and where you're going?21:36
*** ddieterly has quit IRC21:36
mordredjeblair: no. it will make absolutely no problems21:37
jeblairmordred: w00t21:37
SpamapSjeblair: be sure to check the firmware in those. ;)21:37
jeblairmordred: that's the second best conversation i've had all day, after talking with my fishmonger about the electronic frontier foundation.21:37
*** hodos has joined #openstack-infra21:37
*** puranamr has quit IRC21:38
openstackgerritMichal Rostecki proposed openstack-infra/project-config: Determining Rally-Devstack integration on pre_test_hook
jeblairSpamapS: istr hearing how opencompute solved all firmware related issues?21:38
*** ildikov_ has quit IRC21:38
fungijeblair: sounds like you chose wisely among available fishmongers21:39
fungiall the fishmongers here seem to want to talk about college sports instead21:39
*** puranamr has joined #openstack-infra21:39
*** unicell has joined #openstack-infra21:39
SpamapSjeblair: I don't recall hearing that, but point being.. be sure you're not getting somebody else's infected usb/nic/etc. controller. :)21:40
SpamapSseriously thoug21:40
SpamapSWHY did Ubuntu have to break my textcons21:40
fungiSpamapS: when you buy new hardware, chances are you're getting the nsa's infected usb/nic/etc21:41
jeblairSpamapS: here's some unauthenticated information from hackernews!
SpamapSfungi: well we're in merikuh, so thats ok. Its just those non merikuh firmwares.21:41
*** ddieterly has joined #openstack-infra21:42
jeblairi'm sure we could ping some people with nicks starting with J if you want to confirm that21:42
SpamapSjeblair: I've chatted with them before. They've done "a lot"21:43
*** puranamr has quit IRC21:43
SpamapSjeblair: totally worth running things on, but probably also worth _not_ putting the keys to the kingdom on.21:43
SpamapSJun 18 14:41:17 bm-aw1az2-freecloud0001 dnsmasq-tftp[8624]: failed sending /srv/tftp/ubuntu-installer/amd64/linux to
dimsfungi: looking21:44
jeblairSpamapS: how about the load balancers for the git mirrors?21:44
*** unicell has quit IRC21:44
jeblairSpamapS: otoh, super high-profile/public; otoh, not _actually_ our authoritative copy...21:44
SpamapSjeblair: SSL termination... ?21:44
*** spzala has quit IRC21:44
fungiyeah, "back in the day" when i worked somewhere that reused dedicated servers for new customers after root access had been in the control of prior customers, we basically reflashed all the firmware that the vendor's toolkit let you reflash, and hoped that there was no trojan firmware redirecting the firmware reflashing21:44
fungiSpamapS: ssl is passthrough there21:45
funginot being terminated on those21:45
SpamapSthen yeah, probably fine, though.. why onmetal for that? Just for fun?21:46
jeblairSpamapS: probably not, actually.  we just need a load balancer with bandwidth + configurability.  rax lbaas lacks configurability.  rax vms lack bandwidth (we round-robin between 2 vms now for bandwidth reasons, but they lack shared state for backend load)21:46
*** enikanorov2 has joined #openstack-infra21:46
dimsfungi: i like the idea21:46
*** mriedem is now known as mriedem_afk21:46
SpamapSjeblair: oh, vms lack bandwidth?21:46
*** romcheg has quit IRC21:46
*** amitgandhinz has quit IRC21:46
SpamapSseems like.. a kind of important thing for vms to have. ;)21:46
jeblairSpamapS: slightly delayed response (the 'not actually' was about ssl term21:46
*** unicell1 has joined #openstack-infra21:46
*** romcheg has joined #openstack-infra21:46
fungidims: mainly concerned (per my comment there) about the possibility that it could cause problems for any oslo libs still "supporting" 2.6 in their master branches21:47
*** romcheg has quit IRC21:47
*** julim has quit IRC21:47
dimsfungi: true, a quick scan looked ok from that point of view as well21:47
fungidims: great. i'm +2 in that case, once lifeless works out the test_requires workaround21:48
dimsfungi: will keep a watch out when the g-r sync job files reviews21:48
*** _nadya_ has joined #openstack-infra21:49
lifelessI'm thinking we'll make pbr strip the markers for test_requires21:49
lifelessuntil setuptools supports it21:49
lifelessand we'll work on setuptools in parallel21:49
*** tonytan4ever has quit IRC21:49
fungiSpamapS: rackspace rate-limits bandwidth to instances scaled somewhat by the flavor you use21:49
jeblairSpamapS: yeah, we maxed out the 600mbps of an 8g instance a while ago...21:50
jeblairfungi: but it looks like larger vms have more bandwidth?  did that change?  i'm now wondering why we didn't just get a bigger git.o.o vm...?21:50
*** unicell has joined #openstack-infra21:50
*** unicell1 has quit IRC21:51
*** puranamr has joined #openstack-infra21:51
fungijeblair: they do. i think maybe we were trying to be courteous and not take up extra cpu unnecessarily so scaled horizontally instead21:51
SpamapSalso what's lacking in the lbaas?21:51
jeblairSpamapS: inability to tune parameters (such as queue depth) for non-http protocols21:52
SpamapSsince the two answers are "ask for a high bandwidth low CPU flavor" or "ask for lbaas to work"21:52
SpamapSjeblair: oh fail. :(21:52
jeblairSpamapS: (er, well it's lacking the ability. it has the inability.  but you get the idea)21:52
mordredalso - it's not an openstack cloud service21:52
SpamapSis it the old atlas still?21:52
mordredwhich means we'd be tying ourselves a bit more to non-openstack and that would be sadpanda21:52
mordredSpamapS: yah21:53
SpamapSI like what they did with Octavia.. where they made it both its own thing, and a driver for Neutron LBaaS21:53
ianweveryone: the centos jobs are stuck with (stop hpcloud mounting ephemeral disk); appreciate some eyes on it21:53
*** hashar has quit IRC21:53
fungiianw: oh, thanks. i missed you had a change up to get around that21:53
*** bknudson has quit IRC21:54
*** ildikov_ has joined #openstack-infra21:54
dougwigfungi: fwaas change to remove that dependency is going into the gate:
*** jgrimm is now known as zz_jgrimm21:55
fungidougwig: appreciated. abandoned 184384 now21:55
*** _nadya_ has quit IRC21:56
*** tvanderwiel has quit IRC21:56
*** dboik has quit IRC21:57
openstackgerritRamy Asselin proposed openstack-infra/zuul: Add the ability to abort jobs on a patch merge
openstackgerritMerged openstack-infra/project-config: Set some more defaults for Linux Bridge
fungiianw: does 192963 need a corresponding dib element addition?21:58
ianwfungi: i don't think so, because they will never have booted with cloud-init to make the entry?21:59
mordredok. SpamapS patch landed21:59
*** armax has joined #openstack-infra21:59
mordredjeblair, Shrews, fungi, SpamapS: I'm going to cut an 0.8.1 - any objections?21:59
jeblairmordred: you should review ianw's changes21:59
*** dboik has joined #openstack-infra21:59
mordredoh. then I will do that21:59
fungiianw: ahh, yep! you are right21:59
fungii will let mordred look at that one before it gets approved22:00
jeblairmordred: no objections to release22:00
mordredianw: that's so gross22:00
jeblairi thought you'd like that22:00
mordredwell, I mean - STAB22:01
fungimordred: it seems to have started spontaneously after hpcloud began adding a reference to the ephemeral block device in the metadata22:01
greghaynesyeeep, score 1 for no cloud-init on testenvs22:02
mordredmaybe that was supposed to be emcees22:02
fungicloud-init will apparently happily configure you to automatically and persistently mount the ephemeral device on /mnt if it's mentioned in the metadata22:02
ianwyeah, i *think* that changed underneath us -- i mean maybe there's some other explanation22:02
mordredcloud-init is the config management system that is there to bootstrap your config management system but also to configure all of the things because clearly you don't have a config management system22:03
mordredgreghaynes: you haven't happened to have had any time to day to look at why simple-init is having boot problems on rax have you?22:04
mordredgreghaynes: it's next on my thing to battle if nobody else has22:04
*** shardy has quit IRC22:04
greghaynesmordred: nerp, ive been working on a talk22:04
*** trey has joined #openstack-infra22:04
* fungi still needs to do that22:04
ianwso if someone could kick of some rebuilds of images after that merges, that'd be great22:05
jeblairif anyone is looking for a presentation program....
greghaynesaye, I really should have done that22:05
*** Swami has joined #openstack-infra22:05
greghaynesbeen doing slidy.js html all day22:06
*** puranamr has quit IRC22:06
fungijeblair: i'm strongly considering doing that for oscon22:06
* jeblair will hack new slide transitions for beer22:07
*** baoli has quit IRC22:07
jeblairgreghaynes: if you write presentations in rst, you can use pandoc to make slidy (or reveal.js) from the rst.  or just run the rst with presentty.  covers your bases that way.22:08
greghaynesyea, thats definitely a good idea22:08
jeblair(*note: if you use the ..ansi: element it probably won't work in slidy or reveal)22:08
jeblair(but ..image: totally works in presentty ;)22:09
*** dboik has quit IRC22:09
*** dustins has quit IRC22:11
greghaynesmordred: ok, talk done, so ubuntu-minimal + simple-init elements in rax is what failed?22:12
anteayathank goodness presentty has cowsay22:12
anteayait is a must for all my presentation choices22:12
SpamapSTheJulia: fyi, I gave up trying to just use the ubuntu installer. Bifrost here we come.22:13
greghaynesSpamapS: You *really* dont like that node I deployed for you, huh?22:13
SpamapSgreghaynes: its ubuntu's fault here22:13
SpamapSgoing into not-vga mode22:13
*** puranamr has joined #openstack-infra22:13
SpamapSand making it not obvious at all how to do the installer via serial console22:13
*** Swami has quit IRC22:13
greghaynesSpamapS: so youre going to bifrost on the bastion to deploy a node youre going to puppet-ironic?22:14
SpamapSgreghaynes: bingo22:14
mordredgreghaynes: yah22:15
mordredSpamapS: ++22:15
greghaynesmordred: ok, ill poke at that now22:15
mordredSpamapS: I think it's a great idea22:15
SpamapSFor the case where we need to install new bastions.. I'm going to make Ng do it.22:15
greghaynesSpamapS: WFM22:15
*** ddieterly has quit IRC22:16
NgSpamapS: no that's fine, I don't mind at all mounting ISOs from my laptop in .uk to the iLO of servers in .us ;p22:16
*** vnod has joined #openstack-infra22:16
greghaynesNg: glad to hear! ;)22:16
SpamapSNg: did you do that via some kind of vpn onto the ilo network?22:16
SpamapSSee that might be the more sane option... learn to use the evil vpn.22:17
NgSpamapS: yep22:17
Ngif you use the browser version of the iLO jazz, you can attach local files/folders to the BIOS22:17
Ngobviously it requires ActiveX, so I was actually doing it in a VM, attaching to the ISOs on my host OS22:18
Ngbecause why not throw in more layers of hilarilol22:18
SpamapSNg: I've actually succeeded in attaching things via the ssh22:18
SpamapSNg: it does not require activex22:18
SpamapSthe java one does actually work22:18
SpamapSbut only if the network address the ilo has is directly accessible by the java22:19
mordredSpamapS: so, if we have two single-node-ironics, maybe each one has bifrost on it and can blow away the other if needed?22:19
SpamapSit doesn't know how to use socks proxies or SSH forwards like my browser does.22:19
mordredSpamapS: that may be too much for right now22:19
mordredSpamapS: I like your current plan of "Screw it, I'm using bifrost"22:19
SpamapSmordred: this is really a seeding process. We can also just make the second one deployed by the regular bm cloud, and then if we have to redo the first, just convert it.22:19
SpamapSmordred: because we're not trying to do it all from the cloud. We're just getting an OS up.22:20
mordredSpamapS: ++22:21
mordredthat's exactly right22:21
fungi"screw it, i'm using bifrost" should totally be the slogan for bifrost22:21
*** weshay has quit IRC22:21
SpamapSI smell a t-shirt22:21
SpamapSwith Thor on it.22:21
mordreddevananda, TheJulia, NobodyCam: ^^22:21
fungii mean, it is the most convenient route to where we're eventually headed22:22
lifelessmordred: quick q22:22
lifelessmordred: what are your thoughts on pbr growing some significant flexability in the way it achieves its goals, such that pip and virtualenv could use it?22:23
lifelessmordred: you've expressed in the past grave concerns about corner cases.22:23
openstackgerritMerged openstack-infra/project-config: Remove mount of ephemeral disk
openstackgerritMerged openstack-infra/project-config: Remove cloud-init details
mordredlifeless: I would love it if our use of setup.cfg became prevalent22:23
*** __TheDodd__ has joined #openstack-infra22:23
lifelessI don't think it would hurt that goal :)22:23
mordredlifeless: and I think you've aligned us on most of the corner cases22:23
* SpamapS waits patiently for lxml and cryptography to compile.. 22:23
lifelessSpamapS: pip install -U pip22:24
mordredlifeless: I don't want to grow large amounts of super-flexibility22:24
SpamapSlifeless: latest version, first time on this box22:24
lifelessmordred: its almost all aimed at precisely one thing: working from an sdist without pkg_resources or setuptools or pip present at all22:24
openstackgerritRamy Asselin proposed openstack-infra/zuul: Add the ability to abort jobs on a patch merge
lifelessmordred: since thats a hard requirement for them both - e.g. get-pip22:25
mordredlifeless: I believe you have good judgement in this area - my main tentpoles are "have one non-changing way to do most things with baked in defaults" and "git-tag-based-releasing"22:25
mordredlifeless: at this point, as long as that doesn't change, I would LOVE to grow pip and setuptools as clients22:26
mordredor virtualenv22:26
mordredor whatev22:26
*** cbx33 has quit IRC22:26
vnodcan anyone share steps to add channel bot22:27
*** thedodd has quit IRC22:27
lifelessmordred: so, I'll write up a spec and get donald to say it meets his needs, and then you can eyeball it22:27
mordredlifeless: AWESOME22:27
lifelessfungi: the test job that checks a repo's local changes to requirements.txt are compatible with g-r22:28
lifelessfungi: I think we need to virtualenv + pip install that too; remind me where it is ?22:28
lifelessmordred: some of the things are AUTHORS generation and a way to do changelog differently22:29
lifelessmordred: because without gerrit they get muuuuch noiser changes22:29
Clintvnod: which bot?22:29
*** fdegir has joined #openstack-infra22:29
lifelessmordred: I say this now for you to have something to think about22:29
mordredlifeless: yah. well, I think we could do a better job of that too22:29
*** dimtruck is now known as zz_dimtruck22:29
mordredlifeless: esp: re ChangeLog generation bug I mentioned earlier22:30
vnodClint channel bot to cloudpulse22:30
openstackgerritJoshua Harlow proposed openstack/requirements: Mark 'enum34' and 'futures' as only needed on py2.7
NgSpamapS: I think I tried that and got bored of figuring out ports to tunnel ;)22:30
SpamapSNg: exactly22:30
*** e0ne has quit IRC22:31
*** rkukura has quit IRC22:31
greghaynesI used to have a big list for the different types of ilo's of what ports to forward22:31
Clintvnod: look in openstack-infra/system-config22:31
*** rkukura_ has joined #openstack-infra22:31
greghaynesits excruciating, but so is booting a windows vm to run a java applet22:31
anteayaClint: this is the file to allow meetbot in the channel for logging:
SpamapSgreghaynes: and yet, ubuntu developers still insist server users want a graphical console22:33
* SpamapS facepalms22:33
anteayaClint: this is the file to allow infra to access the channel:
anteayaClint: infra access is required for all other channel needs22:33
mordredooh! I just found an easy bug for someone to fix22:33
mordred <--22:33
mordred"The file is hosted in openstack-infra/config"22:33
mordredthat repo has not existed in a while22:33
anteayaClint: this is the file if a project wants patches to be announced in channel:
Clintanteaya: i think you might want to be highlighting vnod22:34
anteayaClint: thought I would share with you22:34
greghaynesSpamapS: Mayble youll just have to grab the alternate install from 12.10, and then upgrade it to trusty ;)22:34
anteayaClint: makes more sense to me to share with you so you know as well22:34
* greghaynes might be having a fun time suggesting bad ideas22:34
Clintanteaya: ah, thanks, i had absorbed that already22:35
openstackgerritlifeless proposed openstack/requirements: Remove hacking and hacking-owned dependencies.
openstackgerritlifeless proposed openstack/requirements: Mark as Python2.7 only the dependencies that are.
openstackgerritlifeless proposed openstack/requirements: Add upper-constraints.txt.
anteayaClint: oh, sorry I missed that in your response22:35
*** EricGonczer_ has joined #openstack-infra22:35
*** eharney has quit IRC22:37
vnodclint anteaya thanks. I will add cloudpulse to openstack-infra/system-config
vnodclint anteaya. Is that the only change needed ?22:38
*** puranamr has quit IRC22:39
mordredvnod: you will need to grant some access, I recommend looking at - it has a section on granting access to the infra bots22:39
mordredspecifically here:
Shrewsmordred: ok on 0.8.122:40
Shrewsif you haven't already22:41
*** EricGonczer_ has quit IRC22:41
*** doug-fish has left #openstack-infra22:41
mordredShrews: just pushed!22:42
openstackgerritDavide Guerri proposed openstack-infra/shade: Centralize exception management for Neutron
mordredlook it's a jhesketh !22:44
*** davidlenwell is now known as david_lenwell22:44
*** david_lenwell is now known as davidlenwell22:44
*** crinkle has quit IRC22:44
mordredjhesketh: so - I JUST NOW pushed a release of shade that should make swift uploads to rax in nodepool happy22:44
mordredjhesketh: it took basicaly all day :)22:44
*** sigmavirus24 is now known as sigmavirus24_awa22:44
vnodmordred sorry if this is basic question22:44
vnodregister channel requires login ?22:45
*** EricGonczer_ has joined #openstack-infra22:45
*** RichardRaseley has quit IRC22:46
*** jasondotstar has quit IRC22:46
*** RichardRaseley has joined #openstack-infra22:46
*** sarob_ has joined #openstack-infra22:47
*** sarob has quit IRC22:48
*** kevinbenton has quit IRC22:48
mordredvnod: yes - well, it requires that you're logged in to nickserv - and it requires you have op on the channel - do you still have op status?22:49
mordredvnod: sdake may have op control on teh channel - if not, we do have some special powers to take it over ... but I'm about to be AFK for a sec22:49
*** TheJulia has quit IRC22:50
vnodsdake setup the IRC,22:50
*** kevinbenton has joined #openstack-infra22:51
vnodso he might have the OP22:51
pc_manteaya: hi22:51
mordredvnod: ok. well, it's not a registered channel, so we may have to take care of it tomorrow22:51
mordredjeblair: ^^ next time you have a few, #openstack-cloudpipe is not registered but should be22:51
vnodI will take it up tomorrow22:52
*** boris-42 has quit IRC22:52
vnodmordred thanks for your help22:52
anteayapc_m: how can I help?22:52
pc_manteaya: Wondering if you know who I could talk to about a question related to keystone authentication?22:52
anteayapc_m: what is the question?22:52
*** yamamoto has joined #openstack-infra22:53
sdakevnod i added you as ops, and had set channel properties correctly for openstack-infra to access22:54
vnodsdake thanks22:54
mordredsdake: woot!22:54
pc_manteaya: I'm wondering is there is a way to do like a proxy authentication. For example, could a tenant give some form of auth token to someone else (ISP/IP) so they could do something on the tenant's behalf (without knowing the tenant's credentials)22:54
sdakehad set a long time ago is what I eman22:54
mordredpc_m: yes22:54
jeblairmordred: i just joined that channel.  there is no one in it.22:54
pc_mmordred: ooooo! tell me more!22:55
mordredjeblair: sorry - cloudpulse22:55
anteayapc_m: well if I followed the keystone meeting yesterday tehre are security concerns around that22:55
sdakecloudpipe lol22:55
mordredI'm sorry22:55
*** openstackgerrit has quit IRC22:55
*** markvoelker_ has quit IRC22:55
*** changbl_ has quit IRC22:55
*** ericksonsantos has quit IRC22:55
*** ianychoi has quit IRC22:55
*** wznoinsk has quit IRC22:55
*** jklare_ has quit IRC22:55
*** mohankumar has quit IRC22:55
*** rlucio has quit IRC22:55
*** cody-somerville- has quit IRC22:55
*** waynr has quit IRC22:55
*** EmilienM has quit IRC22:55
*** loquacities has quit IRC22:55
*** bradm has quit IRC22:55
*** jgriffith has quit IRC22:55
*** hamzy has quit IRC22:55
*** dmsimard has quit IRC22:55
*** jhesketh has quit IRC22:55
* pc_m knew this was the right place to ask22:55
mordredI typed the wrong thing in my acess request22:55
jeblairmordred: it seems to have ops22:55
mordredpc_m: morganfainberg is the person you want to talk to - this is known as "keystone federation"22:55
*** ddieterly has joined #openstack-infra22:55
anteayamorganfainberg: have you any thoughts to share on pc_m's scenario?22:55
mordredpc_m: in the mean time, there is a video from the openstack summit22:56
jeblairmordred: i'm confused; what do you need me to do?22:56
*** angela-s has quit IRC22:56
pc_mmordred: I attended one keystone identity federation session.22:56
mordredjeblair: nothing. I'm sorry - I was wrong - I went to verify chanserv access but I asked for the wrong channel22:56
*** ddieterly has quit IRC22:56
anteayapc_m: good22:56
mordredpc_m: cool. well, morganfainberg is a great contact - although I think he's mostly AFK today22:56
*** sdake_ has joined #openstack-infra22:57
anteayapc_m: does federation not answer your question?22:57
*** yamamoto has quit IRC22:57
jeblairmordred: oh.  ok.22:57
pc_mmordred: It seemed like it was a way to use OAuth to access one cloud from another (if I understood it)22:57
greghaynesmordred: ubuntu-minimal + simple-init is a 1.8GB vhd :(22:57
anteayapc_m: yes22:58
pc_manteaya: Not sure if it will work.22:58
anteayapc_m: do expand22:58
anteayadstanek: are you about?22:58
pc_manteaya: or at least maybe I don't fully understand it well enough...22:58
*** TheJulia has joined #openstack-infra22:58
mordredgreghaynes: yes. it's ludicrous22:58
anteayapc_m: share some details22:58
anteayaseems we are light on keystone folks at the moment22:59
morganfainbergLet me read the scrollback22:59
anteayamorganfainberg: thanks22:59
* morganfainberg is only partially AFK23:00
pc_manteaya: So let's say we have tenant Joe, and Internet Provider Sally. Now, Sally may want to do some things for Joe, say create a subnet, or add a router, without having Joe's credentials.23:00
anteayais Sally the host provider?23:00
anteayaor just providing internet?23:00
pc_manteaya: Let's assume there is some way Joe can give permission to Sally, maybe out of band of OpenStack (like via some token or something).23:00
*** dims_ has joined #openstack-infra23:01
*** sdake has quit IRC23:01
vnodmordred now only change needed is add openstack-cloudpulse to eavesdrop.pp ?23:01
mordredvnod: yup23:01
pc_manteaya: Wondering if there is a way to do that type of scenario?23:01
anteayapc_m: who is Joe's host?23:02
anteayapc_m: Joe is a tenant, Sally is an isp23:02
pleia2StevenK: when you're awake ;) are we good with hitting +A on this or should it still be WIP?
anteayawho is hosting Joe's tenantcy23:02
vnodmordred thanks, will submit the change tomorrow23:02
pleia2StevenK: AJaeger and I are ok with it if you are ready23:02
StevenKpleia2: \o/ at the puppet restart bug fix23:03
pleia2StevenK: yeah, there's still some work to do on the module to make it behave properly, but at least the restarts have ceased23:03
morganfainbergpc_m: is Sally the provider for Joe's tenant? or is Sally a separate (different) provider?23:03
*** dims has quit IRC23:04
pc_mmorganfainberg: Sally would be some IP separate from the cloud.23:04
StevenKpleia2: If you're around to babysit a propsal slave when it lands to just be certain it works even with the Zanata stuff saying it can't generate a zanata.xml, it should be fine to land.23:04
StevenKI fixed the thinko in it last night so it will actually authenticate correctly.23:05
morganfainbergso, what i'm gathering is you want a user (Joe) of your cloud to allow a separate user (external) do some actions for them.23:05
morganfainbergsorry for re-iterating, but tyring to make sure i understand the scenario23:05
pc_mmorganfainberg: So essentially allowing someone (outside of OpenStack) to have access to a tenant to do some (limited) things.23:06
*** puranamr has joined #openstack-infra23:06
*** crinkle has joined #openstack-infra23:06
*** openstackgerrit has joined #openstack-infra23:06
*** mohankumar has joined #openstack-infra23:06
*** markvoelker_ has joined #openstack-infra23:06
*** changbl_ has joined #openstack-infra23:06
*** ericksonsantos has joined #openstack-infra23:06
*** ianychoi has joined #openstack-infra23:06
*** wznoinsk has joined #openstack-infra23:06
*** jklare_ has joined #openstack-infra23:06
*** rlucio has joined #openstack-infra23:06
*** cody-somerville- has joined #openstack-infra23:06
*** waynr has joined #openstack-infra23:06
*** EmilienM has joined #openstack-infra23:06
*** loquacities has joined #openstack-infra23:06
*** bradm has joined #openstack-infra23:06
*** jgriffith has joined #openstack-infra23:06
*** hamzy has joined #openstack-infra23:06
*** dmsimard has joined #openstack-infra23:06
*** jhesketh has joined #openstack-infra23:06
pc_mmorganfainberg: yes23:06
morganfainbergand i am assuming Joe is just a user of the cloud - not the cloud admin23:06
pc_mmorganfainberg: No problem, I'm trying to express the scenario. You have it right.23:06
*** unicell1 has joined #openstack-infra23:06
StevenKpleia2: It will likely need some more polish, but I'm happy for that to occur later -- and having said that, it will certainly need changes to support Horizon, once I figure out how to get zanata to see the PO files in it.23:06
pleia2StevenK: sounds good, I will be around for a bit longer, and can work with AJaeger in the AM if anything really bad goes on (though I think it's unlikely)23:06
pc_mmorganfainberg: And we don't want Sally to have Joe's auth credentials.23:07
pleia2StevenK: have you asked AJaeger about the horizon stuff? he's much more familiar with what goes on there23:07
pc_mmorganfainberg: Hoping Joe can give some kind of token or something.23:07
StevenKpleia2: Yes, I'm quietly confident that I wrote it defensively enough for it to not care if the zanata.xml generation fails.23:07
anteayamorganfainberg: to me this sounded like the security concerns we were discussing the other day23:07
morganfainbergpc_m: give me a second - i need to look at something23:07
*** sdake_ is now known as sdake23:08
morganfainberganteaya: i missed that meeting23:08
pleia2StevenK: yeah, when I was reviewing this morning it eemed ok23:08
*** unicell has quit IRC23:08
pleia2seemed too23:08
StevenKpleia2: I have enough of an understanding based on the horizon repo itself as well as what the current scripts do WRT setup, my stumbling block at the moment is getting zanata to find the bits in the tree.23:08
anteayamorganfainberg: perhaps it was a discussion with you, I can't remember23:08
morganfainberganteaya: doesn;'t sound familiar23:08
anteayamorganfainberg: okay23:08
StevenKpleia2: Which involves XML eating my brain. And perhaps stracing java, I'm not sure.23:08
pleia2StevenK: ah, gotcha23:08
pleia2alright, +A away23:09
morganfainbergpc_m: I *think* you can use either the oauth1.1 (but not 100% sure on that), or the ec2 credential23:09
*** puranamr has quit IRC23:09
StevenKpleia2: It also won't work until lands with the hiera stuff, but I think we went through that before23:10
*** marzif has joined #openstack-infra23:10
pc_mmorganfainberg: For oauth, would Joe do the request and then give the token to Sally?23:11
pleia2StevenK: ah yes, I had lost track of this one23:11
morganfainbergpc_m: so i haven't looked at the real use case of the oauth1.123:11
morganfainbergsince it's oauth1.1 :P23:11
greghaynesmordred: huh, mine booted, then it shut itself off23:12
*** EricGonczer_ has quit IRC23:12
greghaynesmordred: my rax node with ubuntu-minimal + simple-init, that is23:12
*** __TheDodd__ has quit IRC23:12
*** ashleighfarnham has quit IRC23:12
pc_mmorganfainberg: Also, does OpenStack have EC2 credential support?23:12
morganfainbergpc_m: I *think* the ec2 credential is more what you're looking for23:12
* pc_m I need to google on ec223:13
morganfainbergpc_m: the ec2 support in keystone is limited, but you can use the ec2 credential to get a keystone token23:13
pleia2StevenK: ok right, so it won't "work" anyway w/o the zanata.xml so the hiera thing isn't a true dependency here23:14
*** signed8bit has quit IRC23:14
pleia2StevenK: we'll just need to land this once we're ready for this to start chugging along23:14
*** puranamr has joined #openstack-infra23:14
morganfainberggranting arbitrary access to a tenant hasn't been well defined in OpenStack at this point. usually Sally would end up with an account23:14
greghaynesmordred: oh, nvm, its just a borked image23:14
morganfainbergand then would be granted the appropriate role(s) on Joe's tenant (via something like a trust)23:15
pc_mmorganfainberg: And that account would have access to Joe's resources?23:15
*** Longgeek has joined #openstack-infra23:16
morganfainbergpc_m: by resoures you mean log into a VM or perform actions on the OpenStack APIs?23:16
pc_mmorganfainberg: Sally would do things like create a router, set an IP on a GW I/F, create a provider net,...23:17
morganfainbergso even a token would grant access to do things on joe's tenants. but access to the tenant (via an account, etc) - does not grant the ability to login to a vm23:18
*** alexpilotti has quit IRC23:19
pc_mmorganfainberg: yeah no vm access at all needed.23:19
StevenKpleia2: Correct -- with the change landed, it should just log that it can't find the zanata.ini, and then that it can't generate the zanata.xml, and everything else as normal.23:19
morganfainberghow long would sally need access in this case? and how often would this be granted/removed?23:19
*** yamamoto has joined #openstack-infra23:20
morganfainbergpc_m: I'm thinking the OAuth1.1 api meets the needs - and is what you're looking for.23:20
morganfainbergpc_m: i think*23:20
*** otter768 has joined #openstack-infra23:21
pc_mmorganfainberg: Joe would have an enterprise net, that he wants to connect to OS cloud, via some IP. Joe would request Sally to create the connection, and Sally would setup the MPLS connection to the cloud. She would want access to Joe's account to create pieces on the OS side.23:21
pc_mmorganfainberg: Like AWS Direct Connect23:21
morganfainbergok a bit more concrete example helps. thanks23:21
pc_mmorganfainberg: So Sally may want to create a provider net for Joe, set the GW IP address, etc.23:22
pc_mmorganfainberg: To complete the connection. Hence the access. We don't want to give Sally the credentials for Joe, but do want to allow the access (for limited purposes, if needed)23:23
morganfainbergthe Oauth1.1 is the only thing we have that really meets your needs23:23
morganfainbergbut this isn't a workflow we've really had a lot of requests for yet.23:23
*** rlandy has joined #openstack-infra23:23
pc_mmorganfainberg: So Joe would use OAuth1.1 to get a token, and give that to Sally?23:23
openstackgerritMerged openstack-infra/project-config: Support pushing translations to Zanata
morganfainbergpc_m: yes, then sally would use the OAuth workflow to get a Keystone token23:24
*** yamamoto has quit IRC23:24
morganfainbergpc_m: but not everyone enables oauth1 in keystone (until kilo)23:24
morganfainbergmost deployments afaik did not enable it23:24
pc_mmorganfainberg: good to know.23:25
morganfainbergpc_m: i'd like to connect you with stevemar, but he's offline atm23:25
morganfainberghe wrote the oauth1 code and can probably help get you going.23:25
*** otter768 has quit IRC23:25
*** markvoelker_ has quit IRC23:26
pc_mmorganfainberg: That's great. I still need to understand the OAuth1 workflow.23:26
pc_mDo you know stevemar's timezone?23:26
morganfainbergpc_m: he's based in toronto23:26
*** camunoz_gone is now known as camunoz23:26
morganfainbergpc_m: is the documentation on it23:27
pc_mmorganfainberg: Cool. I'm in NH, so can try in the morning23:27
morganfainbergpc_m: come over to #openstack-keystone and look for him tomorrow23:27
pc_mmorganfainberg: Thanks! I'm really shaky on my understanding of OAuth.23:27
morganfainbergsteve is a really good resource for this23:27
*** puranamr has quit IRC23:28
*** boris-42 has joined #openstack-infra23:28
pc_mmorganfainberg: anteaya: Thank you very much!23:28
*** dalgaaf has quit IRC23:28
*** puranamr has joined #openstack-infra23:29
dstanekanteaya: i'm around now. did you need me?23:29
*** whitenoise has quit IRC23:29
anteayadstanek: ha ha ha23:30
anteayapc_m: glad you got it sorted out23:30
anteayadstanek: good timing, morganfainberg just taught us all about oauth123:30
anteayadstanek: thanks though :)23:31
lifelessmorganfainberg: how cruel was that23:32
morganfainberglifeless: uhhhhhhhhh23:32
* morganfainberg hides23:32
pc_manteaya: You going to Neutron mid-cycle?23:32
anteayapc_m: yes, and I see you are as well23:32
*** hichihara has joined #openstack-infra23:32
pc_manteaya: Cool. See you next week!23:32
dstanekanteaya: i was out picking up dinner for the family, but i'm glad i didn't have to torture you23:32
anteayapc_m: I'm looking forward to it23:33
pc_manteaya: Always si fun23:33
anteayadstanek: glad your family is well fed23:33
anteayadstanek: seems the torture is morganfainberg's responsibility23:33
anteayapc_m: 'tis23:33
dstanekPTLs have been known to enjoy it23:34
pc_manteaya: will ttyl, signing out for the night.23:34
*** Longgeek has quit IRC23:34
anteayapc_m: night23:34
*** pc_m has quit IRC23:35
anteayadstanek: have they?23:35
anteayadstanek: I'm losing track23:35
*** puranamr has quit IRC23:35
TheJuliaSpamapS: If there is anything I can do, let me know.  I have a wide open schedule tomorrow :)23:35
*** puranamr has joined #openstack-infra23:36
*** ZZelle_ has quit IRC23:36
*** RichardRaseley has quit IRC23:37
*** EricGonczer_ has joined #openstack-infra23:41
*** EricGonczer_ has quit IRC23:44
*** chlong has joined #openstack-infra23:48
*** RichardRaseley has joined #openstack-infra23:48
*** RichardRaseley has quit IRC23:48
*** rcarrillocruz has quit IRC23:51
*** dannywilson has quit IRC23:51
*** dims_ has quit IRC23:53
*** RichardRaseley has joined #openstack-infra23:54
fungilifeless: if nobody's answered you yet (i haven't skimmed scrollback to see) you want to look at
fungilifeless: it doesn't currently run things from the openstack/requirements repo directly, so may not need a venv?23:56
*** SumitNaiksatam has quit IRC23:56
*** SumitNaiksatam has joined #openstack-infra23:56
openstackgerritJames E. Blair proposed openstack-infra/gear: Do not raise RetryIOError on blocked write
mordredgreghaynes: great! then maybe the nodepool one will also work23:57
*** RichardRaseley has quit IRC23:57
fungilifeless: it just does strict line matches between the project reqs and global reqs23:57
* fungi disappears afk again for a while23:57
mordredgreghaynes, jeblair, fungi: I will, when I wake up, try nodepooling23:57
greghaynesmordred: heh, that was pekac, im got it booting and networking was fail, making one I can actually log in to from console now23:57
greghayneser, that was pebkac23:58
mordredgreghaynes: oh - wait - so it did fail for you23:58
mordredI mean23:58
mordrednot awesome that it failed23:58
mordredbut awesome that it failed, because then fixing is good23:58
mordredgreghaynes: btw - I like your new init-script stuff in dib a lot23:58
*** sarob_ has quit IRC23:59
*** sarob has joined #openstack-infra23:59

Generated by 2.14.0 by Marius Gedminas - find it at!