Wednesday, 2016-03-16

EmilienMdsneddon: I just sent an update on OOO ML about rabbitmq/ipv6 issue00:00
EmilienMit should be fixed by now.00:00
*** xinwu has quit IRC00:00
dsneddonEmilienM, Great news, thanks!00:02
EmilienMdsneddon: can you test again please?00:02
dsneddonEmilienM, Is everything merged to master, or do I need to cherry-pick a patch to test?00:02
EmilienMdsneddon: yeah :00:03
EmilienMand run puppet00:03
EmilienMcurl should work now, please read commit messages to see what I did.00:03
dsneddonEmilienM, OK, I don't know how to pull patches from Github, though, so I'll have to figure that out00:04
*** xinwu has joined #tripleo00:04
dsneddonEmilienM, Wait, I'm confused. Both these patches say "merged to master". Are these not automatically picked up in new Delorean packages or not?00:06
openstackgerritMerged openstack/diskimage-builder: Don't cache debootstrap rootfs by default
*** xinwu has quit IRC00:07
*** xinwu has joined #tripleo00:07
EmilienMdsneddon: sync are done manually in opm00:08
EmilienMdelorean is not yet fully in place yet00:08
dsneddonEmilienM, So are there instructions for testing patches before they merge, or should I wait for OPM packages to be manually built?00:08
EmilienMjayg|g0n3 should build OPM master tomorrow00:09
*** ianw has quit IRC00:10
*** thrash is now known as thrash|g0ne00:12
*** xinwu has quit IRC00:12
*** xinwu has joined #tripleo00:12
EmilienMmichchap: sounds like CI is broken I see your patch failing.
*** xinwu has quit IRC00:20
*** dmacpher has quit IRC00:21
EmilienMlooks like all is failing omg00:22
*** openstack has joined #tripleo00:24
openstackgerritMatthew Thode proposed openstack/diskimage-builder: Add testing for the Gentoo element
openstackgerritDan Prince proposed openstack-infra/tripleo-ci: Metrics tracking for TripleO deployment tasks
*** afazekas has quit IRC00:53
*** afazekas has joined #tripleo00:59
*** shivrao has quit IRC01:06
*** david-lyle has joined #tripleo01:06
*** dmacpher has joined #tripleo01:11
*** afazekas has quit IRC01:25
*** afazekas has joined #tripleo01:28
*** ebalduf has quit IRC01:32
*** panda has quit IRC01:40
*** panda has joined #tripleo01:40
*** shivrao has joined #tripleo01:49
*** yamahata has joined #tripleo01:52
*** mannidi_ has joined #tripleo01:59
*** yamahata has quit IRC02:21
*** yamahata has joined #tripleo02:29
*** shivrao has quit IRC02:34
EmilienMdsneddon: you around? re ipv602:34
*** xinwu has joined #tripleo02:35
*** xinwu has quit IRC02:36
*** TSCHAK_ has joined #tripleo02:39
*** yamahata has quit IRC02:39
*** TSCHAK has quit IRC02:42
*** rhallisey has quit IRC02:45
openstackgerritSwapnil Kulkarni (coolsvap) proposed openstack/tripleo-docs: First documentation for Operational tools
openstackgerritSwapnil Kulkarni (coolsvap) proposed openstack/tripleo-docs: Extend the root device selection documentation
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Test Keystone hooks support - do not merge
openstackgerritSwapnil Kulkarni (coolsvap) proposed openstack/tripleo-docs: Extend the root device selection documentation
*** shivrao has joined #tripleo03:04
openstackgerritSwapnil Kulkarni (coolsvap) proposed openstack/tripleo-docs: Add MidoNet documentation in advanced deployment
openstackgerritSwapnil Kulkarni (coolsvap) proposed openstack/tripleo-docs: Docs for containerized compute node
*** yuanying has quit IRC03:20
*** TSCHAK has joined #tripleo03:25
*** TSCHAK_ has quit IRC03:27
*** morazi has quit IRC03:32
*** xinwu has joined #tripleo03:38
ayoungEmilienM, lets say I wanted to create multiple Rabbit users, say one per compute node,  instead of having all services use the default.  I would want to make a change here:  to add the additional users.03:39
ayoungEmilienM,  seems to be...poorly written?03:40
ayoungwouildn't I use that to create a user?03:40
ayoungrabbitmq_user { 'overcloud-compute-0':03:41
ayoung  admin    => false,03:41
ayoung  password =>  {get_input: overcloud-compute-0-password},03:41
*** ayoung is now known as ayoung_Zzz__04:01
*** yuanying has joined #tripleo04:07
*** links has joined #tripleo04:10
*** ayoung_Zzz__ has quit IRC04:12
*** TSCHAK_ has joined #tripleo04:24
*** TSCHAK has quit IRC04:27
*** akuznetsov has joined #tripleo04:35
*** mannidi_ has quit IRC04:38
*** akuznetsov has quit IRC04:40
*** masco has joined #tripleo04:41
*** akuznetsov has joined #tripleo05:05
*** yamahata has joined #tripleo05:07
*** Goneri has quit IRC05:25
*** oshvartz has quit IRC05:33
*** panda has quit IRC05:40
*** panda has joined #tripleo05:41
*** david-lyle has quit IRC05:45
*** david-lyle has joined #tripleo05:46
*** akuznetsov has quit IRC05:47
*** akuznetsov has joined #tripleo05:48
*** jdob has quit IRC05:50
*** tzumainn has quit IRC05:53
*** rwsu has quit IRC05:58
*** liverpooler has quit IRC05:59
*** ccamacho has quit IRC06:05
*** ccamacho has joined #tripleo06:08
*** aufi has joined #tripleo06:10
*** rwsu has joined #tripleo06:11
*** saneax_AFK is now known as saneax06:17
*** akuznetsov has quit IRC06:22
*** shivrao has quit IRC06:26
*** jtomasek has joined #tripleo06:36
*** shivrao has joined #tripleo06:36
*** dmacpher has quit IRC06:43
*** akuznetsov has joined #tripleo06:45
*** oshvartz has joined #tripleo06:56
*** shivrao_ has joined #tripleo07:02
*** shivrao has quit IRC07:05
*** shivrao_ is now known as shivrao07:05
*** tesseract has joined #tripleo07:16
*** tesseract is now known as Guest6496707:17
openstackgerritSwapnil Kulkarni (coolsvap) proposed openstack/puppet-tripleo: Fix typo
*** jaosorior has joined #tripleo07:18
jaosoriordsneddnon: Still awake?07:18
*** jdob has joined #tripleo07:19
*** Guest64967 has quit IRC07:21
*** shivrao has quit IRC07:23
*** akuznetsov has quit IRC07:23
*** Marga_ has quit IRC07:25
*** leanderthal|afk is now known as leanderthal07:25
*** akuznetsov has joined #tripleo07:33
*** tesseract- has joined #tripleo07:35
*** akuznetsov has quit IRC07:37
*** dshulyak has joined #tripleo07:38
*** Marga_ has joined #tripleo07:38
*** Marga_ has quit IRC07:41
*** Marga_ has joined #tripleo07:41
*** Marga_ has quit IRC07:42
*** chlong has quit IRC07:42
*** Marga_ has joined #tripleo07:43
*** liverpooler has joined #tripleo07:49
*** bvandenh has joined #tripleo07:55
*** mkovacik has joined #tripleo07:57
*** jprovazn has joined #tripleo08:01
*** social has quit IRC08:04
*** liverpooler has quit IRC08:14
*** liverpooler has joined #tripleo08:14
*** paramite has joined #tripleo08:20
*** mikelk has joined #tripleo08:21
*** kbyrne has joined #tripleo08:22
*** Erming_ has quit IRC08:26
*** Erming_ has joined #tripleo08:27
*** yamahata has quit IRC08:33
*** akuznetsov has joined #tripleo08:34
openstackgerritDougal Matthews proposed openstack/tripleo-docs: Deploy and Scale Swift
*** olap has joined #tripleo08:37
*** akuznetsov has quit IRC08:38
*** jpena|off is now known as jpena08:50
*** ohamada has joined #tripleo08:58
*** nijaba has quit IRC09:05
*** social has joined #tripleo09:05
openstackgerritJuan Antonio Osorio Robles proposed openstack/python-tripleoclient: Pass RabbitMQ's user and password from the client
*** ifarkas has joined #tripleo09:07
*** mbound has joined #tripleo09:10
*** nijaba has joined #tripleo09:11
*** nijaba has quit IRC09:11
*** nijaba has joined #tripleo09:11
*** akuznetsov has joined #tripleo09:22
*** jistr has joined #tripleo09:24
*** xinwu has quit IRC09:32
*** derekh has joined #tripleo09:34
*** trown has joined #tripleo09:38
*** lucas-dinner is now known as lucasagomes09:38
*** panda has quit IRC09:40
*** panda has joined #tripleo09:41
*** akuznetsov has quit IRC09:41
jaosoriorjistr: as part of the update scripts is rabbitmq restarted too?09:41
jistrjaosorior: yeah on stack-update, currently rabbitmq should be restarted too09:42
jaosoriordo I have to do something specific to do that?09:42
*** dtantsur|afk is now known as dtantsur09:42
jaosoriorI tried changing the user and password for rabbitmq in my overcloud and it all went boom09:43
jaosoriorhad to restart rabbit manually from pacemaker09:43
jaosoriorjistr ^^09:45
*** mannidi has joined #tripleo09:46
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Remove defaults for the RabbitMQ credentials
*** openstackgerrit has quit IRC09:53
*** openstackgerrit_ is now known as openstackgerrit09:53
*** openstackgerrit has quit IRC09:53
*** openstackgerrit_ has joined #tripleo09:53
*** openstackgerrit_ is now known as openstackgerrit09:54
*** openstackgerrit_ has joined #tripleo09:55
*** openstackgerrit has quit IRC09:55
*** openstackgerrit has joined #tripleo09:56
*** paramite has quit IRC09:56
openstackgerritSagi Shnaidman proposed openstack/instack-undercloud: Use installed tempest version
*** gfidente has joined #tripleo10:13
*** gfidente has quit IRC10:13
*** gfidente has joined #tripleo10:13
*** electrofelix has joined #tripleo10:15
*** rhallisey has joined #tripleo10:31
*** ohamada has quit IRC10:31
*** ohamada_ has joined #tripleo10:31
rascathrash|g0ne, jistr, using repos from --repo-setup I was able to introspect the nodes, but as you can see from here for some reason the overcloud deploy script seems to not be able to use any of the introspected nodes, even if they have the tag "baremetal"10:34
rascaany suggestion?10:34
chemhe, I've updated/fixed the patch from EmilienM to rabbitmq:
chemthis should work ok now10:39
*** leanderthal has quit IRC10:39
*** mgould has joined #tripleo10:40
chemjpena: if you still have this ipv6 env around could you test it with the above patch ?10:41
jpenacheM: i returned it an hour ago, but I can try go get a new one10:41
jistrrasca: dtantsur would you have any idea what could be wrong in rasca's case? it's strange that the state of the machines is `power on`... if they're not used, they should be `power off` i think. rasca are they still `power on` now?10:41
chemjpena: I'm going to try on a an simple rabbitmq host just to check10:42
rascajistr, dtantsur yes, they are10:42
rascajistr, dtantsur I can force off them if it helps10:42
dtantsurrasca, jistr, all nodes are in "manageable" state, something probably went wrong10:43
dtantsurrasca, are you sure that introspection has finished?10:43
*** mannidi_ has joined #tripleo10:44
rascadtantsur, jistr since the output is this one even if it says completed, then I suppose as you are saying that something went wrong10:45
*** akrivoka has joined #tripleo10:46
dtantsur"Introspection didn't finish for nodes 018615be-c97b-4961-974b-eb4b74c94500,1adc844a-0f30-4f81-aa42-746f72f403ca,a3462a84-cb93-42b7-aad0-d01187a0dc15,f1a39823-c9b0-470a-8260-8e2cb855bdb2"10:46
dtantsurhmm hmm10:46
*** mannidi has quit IRC10:48
*** paramite has joined #tripleo10:53
jaosoriorgfidente: Hey dude, got time to check this out?
rascadtantsur, I'm trying to run it again10:57
rascajust in case10:57
derekhadarazs: the manaul job I kicked off last night failed due to some error installing a crypto package , I'm kicking it off again10:59
adarazsderekh: thanks!10:59
derekhadarazs: send me a public key and I'll add it to the instance10:59
adarazsderekh: btw can you show me how you do this reproducing?10:59
derekhadarazs: ssh fedora@
*** tosky has joined #tripleo11:01
derekhadarazs: I can't should you as only 3 of us have that access to the cloud to do it11:01
derekhadarazs: sorry, use jenkins user11:02
derekhadarazs: not fedora11:02
adarazsderekh: just wanted to say it doesn't work :)11:02
derekhadarazs: then screen -x , switch to tab 3, thats where the ci command is running11:02
derekhadarazs: tab 4 is on the undercloud, but that only being set up at the moment, it should be an actual undercloud in 20 minutes or so11:03
adarazsderekh: ack, I'm there. thanks11:04
derekhNo failed jobs since midnight, crazy
chemjpena: just in case I have updated the code to include all the functions so the patch from stdlib is not required (it brokes the standard case where you do not have puppetlab-stdlib > 4.11.0 - which doesn't exist yet)11:05
jpenachem: ack. Just to be sure, I should test this against master, right?11:07
rascadtantsur, well I think introspection has got some problems11:08
rascadtantsur, it takes too long to complete11:08
dtantsurrasca, maybe something here will help:
*** trown is now known as trown|relocate11:12
rascadtantsur, something seems broke here11:13
rascadtantsur, note that with ospd7 or 8 introspection is working11:14
*** openstackgerrit has quit IRC11:18
*** openstackgerrit has joined #tripleo11:18
chemjpena: I think it won't works, the management service is listening on the ipv4 address the ip_node_address is not taken into account for management service11:18
chemjpena: the curl code is fixed, but the management service is still listening on a ipv4 address (as it has always did)11:19
jpenachem EmilienM: wasn't that fixed in RHEL?11:19
chemjpena: dunno, I'm using full upstream code here, do you have a pointer ?11:20
chemjpena:on a simple rabbitmq one node install11:20
jpenachem: let me check, I think it was something in the rabbitmq package itself11:20
chemjpena: that would be sweet :)11:22
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
jpenachem: bug 1299214 in rabbitmq-server "Cannot deploy rabbit in an ipv6 environment" [Urgent,Closed: errata] - Assigned to jeckersb11:23
chemjpena: not sure, have to be tested, here the rabbitmq server is properly listening on ipv6 but it's only the management server, port 15672 which is still listening on ipv411:25
jpenachem: that was part of the problem. The management server didn't listen on ipv6, so the cluster didn't form11:25
chemjpena: oki, do you know where I can pull the right version of the rabbitmq server then ?11:26
*** bvandenh has quit IRC11:34
*** weshay has joined #tripleo11:34
*** olap has quit IRC11:34
*** dtantsur is now known as dtantsur|brb11:34
*** olap has joined #tripleo11:36
*** dshulyak has quit IRC11:46
*** veteran has joined #tripleo11:49
toskyon RDO Mitaka snapshot from last week I tried to run Sahara scenario tests with the demo tenant; Sahara uses heat, and demo seems to be configured properly, but then in the logs I see11:52
toskyForbidden: You are not authorized to perform the requested action: identity:list_roles (HTTP 403)11:52
toskyfull log:
toskyis it something that should have been set by TripleO?11:53
EmilienMderekh: failures are related to puppet or not?11:54
EmilienMderekh: we changed something in puppetlabs-rabbitmq11:54
EmilienMbut should not break11:54
derekhEmilienM: which failures?11:54
EmilienMderekh: tripleo CI was failing last night but before my work on puppet rabbitmq11:55
derekhEmilienM: they were all failing trying to install python2-pyasn1 , don't know why but its ok again now11:56
EmilienMderekh: ok :)11:56
EmilienMsorry for noise11:56
derekhEmilienM: np11:56
*** dtantsur|brb is now known as dtantsur12:01
adarazsderekh: hm, so the run failed. what's weird is that the OS_AUTH_URL is not IPv6. I think that should be ipv6 on normal full ipv6 deployment. at least it was on ospd7, not sure how it was implemented in tripleo now.12:06
*** jaosorior has quit IRC12:08
derekhadarazs: are you able to confirm It actually tried to do a IPv6 installation, i.e. did a screw something up?12:09
adarazsderekh: looking.12:09
adarazsderekh: oh wait, never mind, that last printout was from the stackrc, that is supposed to have an IPv4 auth_url12:10
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Set default locale to image in ubuntu-minimal
*** jaosorior has joined #tripleo12:11
adarazsderekh: but the error seems to be "/tmp/tripleo-ci/scripts/ line 112: /home/stack/overcloudrc: No such file or directory" -- maybe creating the overcloudrc somehow fails when the addresses are ipv6, I'll look into the script.12:11
adarazsI'll just go out to grab some lunch before.12:11
derekhadarazs: I can't see logs for the puppet run, Keepalived_vrrp is going crazy on the overcloud controller and the logsd have rotaded out12:11
*** adarazs is now known as adarazs_lunch12:11
derekhadarazs_lunch: ok, don't worry about the overcloudrc error, this is the real error12:12
derekhStack overcloud CREATE_FAILED                                                                               |12:12
derekhthe rest is what happen when the ci job tried to collect logs12:12
adarazs_lunchderekh: you're right. I assumed it will pass because it passed in the gate job12:12
*** olap has quit IRC12:13
adarazs_lunchderekh: hm, it died at step 2 with code 6, that's more consistent with the non-pacemaker job. which gate did you recreate? the nonha or the ceph?12:13
derekhadarazs_lunch: anyways grab lunch, I think what we need to find out is why Keepalived_vrrp is looping12:13
adarazs_lunchderekh: the ceph (where I didn't enable pacemaker) has a known bug with rabbitmq.12:14
adarazs_lunchthe ha and nonha should be the ones to look at while EmilienM is trying to fix the puppet stuff for the non-pacemaker case12:14
derekhadarazs_lunch: ok, that was a ceph job, gonna kick of the nonha one12:15
*** olap has joined #tripleo12:15
adarazs_lunchderekh: thanks!12:15
*** miles has joined #tripleo12:15
*** jayg|g0n3 is now known as jayg12:17
*** mgould has quit IRC12:18
openstackgerritChristopher Brown proposed openstack/os-net-config: Use interface operstate to determine nic status
toskyifarkas: ping generated deployer-input file, who is writing it, or better, which component should I file a bug against if I need to extend the information written there? (Sahara plugin list)12:22
ifarkastosky, tripleoclient12:23
toskyifarkas: thanks12:23
*** rain has joined #tripleo12:24
*** rain is now known as Guest9775812:24
*** Guest97758 is now known as leanderthal12:24
toskyifarkas: I see that the values there are hardcoded, wouldn't it make sense to move it to a file? Or is it like that to allow future expansions with values that are dynamically computed?12:26
*** rodrigods has quit IRC12:26
*** rodrigods has joined #tripleo12:26
*** thrash|g0ne is now known as thrash12:27
ifarkastosky, yeah, that's done this exactly for that reason. we might need to compute more stuff12:27
*** colonwq has joined #tripleo12:28
*** pcaruana has quit IRC12:28
toskyifarkas: so, I'd like to have the set of default sahara plugins (which depends on the version) set there, but I'm not sure how to compute that;12:29
toskyifarkas: given that the set is stable per-version, and the release is close, can I just submit a review with an hardcoded list?12:29
ifarkastosky, yeah, I think the only thing you need to check is if sahara is enabled or not12:31
ifarkasbut a hardcoded list if fine12:31
toskyifarkas: but I don't need to do that, the autodiscovery takes care of it12:31
ifarkastosky, the you don't need to touch the deployer-info file at all, right?12:32
toskyifarkas: the autodiscovery doesn't cover that12:33
toskyifarkas: let's put it this way: if that list of plugin is set, and sahara = False, the tests for Sahara are skipped anyway, so no harm12:33
toskyif the list is set, and sahara = True, the tests have the proper input12:33
ifarkastosky, yeah, that makes sense12:34
jaosoriord0ugal: Are you around?12:34
*** trown|relocate is now known as trown12:35
*** lucasagomes is now known as lucas-hungry12:35
d0ugaljaosorior: Yup!12:39
*** thrash has quit IRC12:40
*** rdopiera has joined #tripleo12:40
*** dshulyak has joined #tripleo12:43
*** ramishra has quit IRC12:43
*** rabi has joined #tripleo12:44
*** thrash has joined #tripleo12:44
*** thrash has joined #tripleo12:44
jaosoriord0ugal: Hey dude, I remember you reviewing some python-tripleoclient code, got some time to check this out?
jaosoriord0ugal: For a strange reason I cannot add you directly to a review :/12:44
*** adarazs_lunch is now known as adarazs12:44
d0ugaljaosorior: My gerrit account is slightly broken, that hasn't worked for some time now.12:44
d0ugalbut yeah, I can do a review now.12:44
dtantsurshadower, hi! please see my responses on
d0ugaljaosorior: Have you seen the most recent comment? Sounds like there might be an issue with it?12:46
*** rabi is now known as ramishra12:46
jaosoriord0ugal: Hadn't seen it12:47
d0ugaljaosorior: Also, that is t-h-t not tripleoclient, I am less useful on that repo :)12:47
d0ugaljaosorior: Did you mean to link this one?
thrashjaosorior: d0ugal that comment on the tht patch is probably because they did not have the tripleoclient patch12:48
jaosoriord0ugal: haha sorry, I meant to pass you the tripleoclient patch12:48
*** dweaver has joined #tripleo12:48
jaosoriorthrash: And that is the case12:48
d0ugaljaosorior, thrash - k, I'll review
thrashjaosorior: quick comment tho... I don't like the username being mixed in with the passwords.12:48
jaosoriorthrash: I don't either, but it's the easier way I found to do it. Any other ideas?12:48
thrashjaosorior: for one thing, it's going to get generated as a long hash.12:49
jaosoriorthrash: That's already the case for the rabbitmq userid of the undercloud12:49
thrashjaosorior: if it isn't specified already in the conf file.12:49
thrashjaosorior: is that really something that we want for the overcloud too?12:49
*** masco has quit IRC12:50
d0ugalIf we renamed the password generation to random string generation it would make sense :)12:50
jaosoriorthrash: I could add an option to specify the user for rabbitmq12:50
*** pradk has joined #tripleo12:51
jaosoriord0ugal: or random credential generation :P12:52
d0ugalheh, true12:52
thrashjaosorior: I may be out of bounds here, but I think a sane default for the user should be in the templates.12:52
d0ugalthrash: I think I agree, we want less in the CLI, not more.12:52
thrashI'm certainly ok with the password, as that's a security thing.12:53
jaosoriorthrash, d0ugal: Would it be OK then if I leave the default user, but leave the password there?12:53
*** Goneri has joined #tripleo12:53
thrashjaosorior: that's what I'm thinking.12:53
*** jprovazn has quit IRC12:53
jaosoriorthrash, d0ugal: Alright, lets do that then12:53
thrashjaosorior: ack12:53
thrashd0ugal: ci passed. +2 it quick!12:53
thrashd0ugal: and a backport:
*** jprovazn has joined #tripleo12:55
d0ugalthrash: done :)12:55
openstackgerritJuan Antonio Osorio Robles proposed openstack/python-tripleoclient: Pass RabbitMQ's password from the client
shadowerdtantsur: I'll have a look12:57
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Remove default for the RabbitMQ password
*** Goneri has quit IRC12:59
*** Goneri has joined #tripleo13:00
*** tzumainn has joined #tripleo13:00
*** tzumainn has quit IRC13:00
*** tzumainn has joined #tripleo13:01
*** dprince has joined #tripleo13:02
*** rcernin has joined #tripleo13:03
*** morazi has joined #tripleo13:03
*** rbrady has joined #tripleo13:06
adarazsderekh: do you know why the job stopped here after --overcloud-deploy/13:08
derekhMust be a problem with OVERCLOUD_DEPLOY_ARGS , one of the files might be missing13:09
derekhadarazs: ^13:09
openstackgerritMichele Baldessari proposed openstack/python-tripleoclient: Clarify error when a template is missing
adarazsderekh: that net-iso.yaml should be copied there by one of the scripts. I think deploy.sh13:10
derekhadarazs: checking on tasb 413:11
adarazsderekh: this line should copy it13:12
chemEmilienM: the other patch to puppetlab-rabitmq required to have this stuff working
chemjpena ^13:14
openstackgerritMerged openstack/python-tripleoclient: Add 'undercloud upgrade' command
*** akuznetsov has joined #tripleo13:16
derekhadarazs: that line is for another script, this is the line that copies the whole tripleo-ci repo into /tmp on the undercloud13:17
*** morazi has quit IRC13:17
derekhadarazs: I think the problem is that the patch us wrong13:17
derekhadarazs: should be /tmp/tripleo-ci/test-environments/net-iso.yaml13:17
*** pino|work has quit IRC13:17
derekhadarazs: will update and kick it off again13:18
*** mannidi_ has quit IRC13:18
EmilienMchem: ok, reviewing 445 first13:18
*** jdob has quit IRC13:18
*** bvandenh has joined #tripleo13:19
*** jdob has joined #tripleo13:19
chemEmilienM: ack13:19
*** lblanchard has joined #tripleo13:19
adarazsderekh: hmmm, yeah, this has been recently changed :/13:19
adarazsderekh: it was the right path until a few days ago13:20
adarazsI'll update the change13:21
derekhadarazs: yup, ok its off again, bbiab13:21
chemEmilienM: updated 44413:22
EmilienMchem: 445 merged13:23
*** pino|work has joined #tripleo13:23
chemEmilienM: nice, reading erb with 3 levels of if is always a pleasure :)13:25
EmilienMchem: 444 merged13:25
EmilienMkicking a OOO job now to check if nothing broke13:25
chemEmilienM: checking in my little set up as well13:25
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: CI test - never merge
*** lucas-hungry is now known as lucasagomes13:26
EmilienMchem: triple checking in puppet openstack CI:
openstackgerritOpenStack Proposal Bot proposed openstack/diskimage-builder: Updated from global requirements
openstackgerritAttila Darazs proposed openstack-infra/tripleo-ci: WIP: Use IPv6 on the gate jobs a test
chemEmilienM: ack13:27
EmilienMadarazs: all rabbitmq patches are merged now13:27
EmilienMadarazs: let us know if you still hit the issue13:28
adarazsEmilienM: i just pushed a change to the CI job, so we'll probably test with those new changes.13:28
adarazsderekh: the -updates job is the same as the -ceph was before?13:29
*** morazi has joined #tripleo13:30
*** jdob has quit IRC13:31
EmilienMadarazs: it should be13:31
*** jdob has joined #tripleo13:31
EmilienMadarazs: but derekh can confirm13:31
EmilienMI'm aware there is some WIP to test ceph again13:32
openstackgerritOpenStack Proposal Bot proposed openstack/python-tripleoclient: Updated from global requirements
adarazsyeah, I'm trying to make sure that the changes to the ceph part are actually ran:
chemEmilienM: jpena on my little rabbitmq one node test, the rabbitmq_management server is listening on ipv6 and the curl command succeeded using latest git from puppetlab-rabbitmq and puppetlab-stdlib13:33
adarazs(line 54-57)13:33
EmilienMchem: !!!13:33
jpenachem: cool! What happened with the ports (15672 vs 25672)?13:33
chemjpena oki, that's the last unknown, I don't know what is listening on the 2567213:34
chemjpena but the 15672 is the management server and it's listening on ::113:34
chemjpena (i specified ::1 for node_ip_address)13:34
jpenachem: ok, that's fine then I guess13:35
*** d0ugal has quit IRC13:35
*** d0ugal has joined #tripleo13:35
*** MaxPC has joined #tripleo13:35
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add neutron profiles
michchapEmilienM: ^ haven't done HA ones yet.13:37
EmilienMmichchap: excellent. What is the progress on keystone?13:37
michchapEmilienM: I'm going to add param doc which should re-kick verify13:38
EmilienMmichchap: other jobs are passing?13:38
michchapEmilienM: is tripleo CI OK now?13:39
EmilienMmichchap: yeah should be, see ML13:39
chemjpena: the 25672 port seems to be related to clustering : grep this
EmilienMmichchap: and see
michchapEmilienM: ok. The first error on HA was because I messed up db sync but the second one I think was ci breakage13:40
EmilienMmichchap: cool13:40
EmilienMmichchap: I'll review it today. And follow-up with Glance, ok?13:40
*** bvandenh has quit IRC13:40
*** panda has quit IRC13:40
michchapEmilienM: yep, cheers13:40
chemjpena: hummm ... but now it's listening on ipv413:40
chemjpena: that's insane13:41
*** panda has joined #tripleo13:41
jpenachem: can you send me the bits you are testing, so we can check in parallel?13:41
chemjpena : I'm using this simple manifest
*** shardy has joined #tripleo13:42
chemon a vm with the puppetlabs-rabbitmq and puppetlabs-stdlib comming from git everything else pulled from a puppet module install13:43
jpenachem: ok, let me give that a try13:43
jaosoriorthrash: Just verified that the new changes worked in both ha and non-ha. Can you check again ?13:48
jaosoriord0ugal: Thanks for the review dude13:48
thrashjaosorior: sure thing.13:48
d0ugaljaosorior: np13:48
jpenachem: did you get voxpopuli/staging from puppet module install or git?13:48
chemjpena module install13:49
*** athomas has quit IRC13:51
thrashjaosorior: just a nit, but shouldn't the client change depend on the tht change, instead of the other way around?13:52
thrashd0ugal: what do you think? ^^13:52
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add keystone and db sync profiles
EmilienMmichchap: excellent work on
jaosoriorthrash: the client change doesn't depend on the t-h-t change. As it can easily work without it. meaning, It'll just overwrite the default value into whatever's given13:53
thrashjaosorior: my thinking is that the tht change can be used without the client changes13:53
jaosoriorthrash: On the other hand, the the t-h-t change cannot be used without the client change. Because it would fail saying that the password wasn't given13:54
michchapEmilienM: Thanks. HA is going to be painful I think13:54
jpenachem: I've tested that. About port 25672 listening on IPv4, I was missing RABBITMQ_SERVER_START_ARGS="-proto_dist inet6_tcp" in rabbitmq-env.conf, iirc that's something we can add to the manifest13:54
* jpena goes for lunch, will be back in ~1h13:54
EmilienMmichchap: maybe with a deps.pp ?13:54
*** jpena is now known as jpena|lunch13:54
thrashjaosorior: i guess it just depends on how you view it. Obviously we both view it the opposite way. :D13:54
thrashlike I said, just a nit. :)13:54
michchapEmilienM: deps.pp?13:54
EmilienMchem: btw, I moved the dep from the fork to voxpopuli yesterday13:55
chemjpena ack, you mean in the upstream puppetlab-rabbitmq module, don't you ?13:55
*** jprovazn has quit IRC13:55
jaosoriorthrash: Well, thing is, if the t-h-t change doesn't have the depends-on pointing to the client change, it will fail because the rabbitpassword no longer has a default (which is now given by the client)13:55
chemEmilienM: Yep, I've seen :)13:55
thrashjaosorior: good point.13:55
EmilienMmichchap: something like in puppet-nova13:55
thrashjaosorior: carry on.13:55
jaosoriorActually ccamacho tried it, and had gotten an error when he tested it13:55
chemEmilienM: so you've seen the comment from jpena: we need to have RABBITMQ_SERVER_START_ARGS="-proto_dist inet6_tcp" for the cluster service to be listening on ipv613:56
EmilienMmichchap: can be useful if you need deps for pacemaker13:56
jaosoriorby the way, ccamacho, thanks for testing the changes, very well appreciated13:56
*** athomas has joined #tripleo13:56
michchapEmilienM: like an anchor set?13:56
EmilienMchem: yeah, don't we set it in tripleo alreaedy?13:56
d0ugalthrash, jaosorior - let's make them depend on each other :P13:56
chemEmilienM: jpena dunno13:56
EmilienMchem: but yeah it should be default in puppetlabs-rabbitmq13:56
thrashd0ugal: perfect!13:56
* d0ugal isn't serious13:56
jaosoriord0ugal haha I actually needed something like that not too long ago13:56
chemEmilienM: jpena how you do that ? (modifying a conf file hold by puppet ?)13:57
jaosoriorI wonder if it's possible...13:57
chemEmilienM: thanks :)13:57
EmilienMmichchap: yeah, maybe... but we first need feature parity13:57
d0ugaljaosorior: I don't think so13:57
michchapEmilienM: OK yep I get it. Wouldn't that already be provided by the neutron module though?13:57
*** bvandenh has joined #tripleo13:57
jaosoriorwould be a nice feature13:57
michchapEmilienM: I agree that's useful.13:57
EmilienMmichchap: right but nothing about pacemaker.13:57
EmilienMmichchap: well, that would be a future step13:57
EmilienMonce all is moved under puppet-tripleo13:57
michchapEmilienM: gotcha, so we put some anchors in for pacemaker resources in the HA profiles13:57
michchapEmilienM: I like that idea13:58
EmilienMmichchap: first step is to steal code from THT so we can manage puppet code on puppet-tripleo13:58
d0ugaljaosorior, thrash:
* chem goes for lunch as well13:58
ccamachojaosorior, but after using the correct patch, the overcloud have deployed correctly, yw13:58
thrashd0ugal: haha13:59
thrashd0ugal: point being, it's actually happened. :P13:59
*** jprovazn has joined #tripleo13:59
d0ugalthrash: Yeah, seems so :) and probably killed CI when it did13:59
openstackgerritGiulio Fidente proposed openstack-infra/tripleo-ci: Enable ceph and netiso for the nonha job
jaosoriorthrash, d0ugal: Well,  I will probably need something like that for this series: doing without some circular dependency... or something is the sort is gonna be painful14:01
thrashjaosorior: yeah... good luck with that. :D14:01
d0ugaljaosorior: hah, good luck14:01
* d0ugal walks slowly away14:01
* thrash backs away quickly14:01
*** bvandenh has quit IRC14:04
*** mgagne has quit IRC14:07
*** mgagne has joined #tripleo14:07
derekhEmilienM: adarazs: the upgrades job is currently just a nonha job with no introspection,  this patch is intended to turn ceph on for non ha
derekhI see I got some typos to fix on that, fixing now14:09
derekhEmilienM: adarazs then there is another patch to turn upgrades on for the upgrades patch14:09
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: compute: include VIR_MIGRATE_TUNNELLED when doing VM shared storage
*** absubram has joined #tripleo14:11
EmilienMgfidente: ^14:11
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Deploy ceph in the nonha job
derekhgfidente: ^14:11
adarazshm, okay. I need to change my patch to have a job that doesn't have pacemaker then. I added pacemaker to the nonha to test that better. but now I don't have anything to test the non-pacemaker case.14:11
gfidentederekh, hey I didn't see that thanks!14:11
*** absubram_ has joined #tripleo14:12
derekhgfidente: np14:12
gfidentederekh, so we don't enable netiso in nonha yet ?14:12
derekhgfidente: correct, should we?14:13
*** absubram has quit IRC14:15
*** absubram_ is now known as absubram14:15
EmilienMchem: the rabbitmq patches work for me14:16
gfidentederekh, I am not sure, you're thinking to use ha for pcmk(3)/netiso/v4 ; upgrades for pcmk(1)/ceph/netiso/v6 and nonha for non-pcmk/ceph ?14:22
derekhgfidente: looks like its heading that way, we should do up a matrix, things have been evolving a lot recently, a target would be nice14:25
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add keystone and db sync profiles
gfidentederekh, actually I'm +2 because I figured that having at least one job which does not use netiso is useful as it covers a lot of "code" in the templates which wouldn't be excercised14:28
toskyifarkas: ok, I was wrong: the set of plugins _is_ configured in puppet (even if hardcoded for now), is there an easy way to retrieve it to have it dynamically set in deploy_input?14:29
gfidenteadarazs, ^^ so I think our target job for ipv6 should probably be "upgrades" ?14:29
toskyifarkas: and by configured (but not tunable right now) I mean this
adarazsgfidente, derekh: where's "upgrades"? I don't see it defined here: -- isn't it using a regular job like nonha? if yes, how do you use ipv6 in only that job?14:31
gfidenteadarazs, you can add a case for it, like it happens for ha/nonha/containers14:32
ifarkastosky, sorry, I don't know14:33
toskyifarkas: np, who should I ask to? :)14:33
*** ayoung has joined #tripleo14:33
derekhadarazs: its not there yet, the upgrades job is running and just doing a normal nonha ci job, we should be adding it soon14:33
adarazsgfidente: case for it where? can you point me to the right part of the ci code?14:33
adarazsderekh: hm, okay. so the name of it is a lie. :)14:34
gfidenteadarazs, an actual bash case I mean
derekhadarazs: correct, thats what gfidente was doing here
derekhadarazs: but its not in yet14:34
gfidenteadarazs, derekh actually let me update that too14:35
adarazsoh, things are moving fast :)14:35
toskyryansb, zaneb: hello heat experts, I have an issue with Sahara and its Heat engine on RDO Mitaka (snapshot from last week):14:37
ifarkastosky, maybe d0ugal or shardy ?14:38
toskyifarkas: oki, thanks; after the other issue I have now14:38
toskyryansb, zaneb: I tried to run Sahara scenario tests with the demo user/project; Sahara provisioning fails and in the heat logs I see:14:38
ifarkasah, ok :-)14:38
toskyryansb, zaneb: Forbidden: You are not authorized to perform the requested action: identity:list_roles (HTTP 403)
toskyryansb, zaneb: is it something that should have been set by TripleO?14:39
toskyactually, it involves also keystone, so the question is also for ayoung ^^14:39
ryansbthe likelihood of that being a Heat problem isn't very high, looks like it's just surfacing the fact that your user can't do that action14:39
*** saneax is now known as saneax_AFK14:40
*** absubram has quit IRC14:40
*** chem has quit IRC14:41
toskyryansb: what permission should the demo user have, apart from heat_stack_owner?14:41
ayoungtosky, look at the policy file on keystone14:42
ayoungit is in /etc/keystone/policy.json14:42
* tosky looks14:43
*** absubram has joined #tripleo14:43
toskyit is "identity:list_roles": "rule:admin_required",14:43
ayoungtosky, is that not unfortunate?14:43
ayoungis that not absurd?14:43
ayoungis that not the band of my exist....ok maybe overstating a little14:44
toskyprobably - then I need to understand what changed compared to a liberty installation where it works14:44
ayoungtosky, but...what needs to list roles?14:44
toskywhere the permissions are the same14:44
ayoungtosky, it was probably being performed by an admin user14:44
toskyayoung: no, I run the tests from a "demo" user14:45
toskyI'm sure that didn't change14:45
ayoungmaybe shardy can confirm, but I suspect now we are using trusts where before it was all done as the heat admin?14:45
ayoungtosky, it was not you14:45
ayoungit was, I suspect, what heat is using to set up delegation14:45
ayoungcuz nothing else changed14:45
ayoungtosky, that rule for list_role goes back to Diablo14:46
ayoungtosky, how are you triggereing this error? what is the code path (in general terms)?14:48
toskyayoung: I don't know that code too much (as QE, it's quite a black box), but let me check14:48
openstackgerritGiulio Fidente proposed openstack-infra/tripleo-ci: Use --overcloud-update in the upgrades job
adarazsderekh: we got the expected "Service Unavailable (HTTP 503)" error in the test deploymet14:49
toskyI can tell you that Sahara uses Heat to orchestrate the creation of an Hadoop/Spark/Storm cluster14:49
toskyso it creates Heat stacks and passes them to Heat14:50
*** veteran has quit IRC14:50
adarazsderekh: the endpoints look okay... maybe we should check if there's anything sensible in the glance logs on the controller?14:51
adarazsderekh: though in the gate job there was nothiing14:51
derekhadarazs: yup, poking around now14:53
toskyayoung: is there an easy way to set the old way of authorization (trusts) to see if it is the issue?14:54
openstackgerritJuan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Remove default for the RabbitMQ password
jaosoriorbnemec: Done ^^14:54
ayoungtosky, yeah, it is a heat config issue.  Let me see if I can find it14:55
*** akuznetsov has quit IRC14:55
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Properly setup DNS for the container CI job
derekhadarazs: see me on tab4, image create worked14:56
adarazsderekh: I see.14:56
*** jprovazn has quit IRC14:56
adarazsderekh: okay, so really no idea why glance gives us 503 during the job14:56
ayoungtosky, I think it is14:57
ayoungunset stack_user_domain_name14:57
derekhadarazs: looks like it might be the proxy14:58
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
adarazsderekh: hm, why is the job using a proxy?14:58
adarazsderekh: probably it can't proxy the ipv6 traffic14:58
toskyayoung: it is already not set in /etc/heat/heat.conf on the controller14:58
derekhadarazs: for the IPv4 we set noproxy for the enpoints14:59
derekhadarazs: 2016-03-16 13:53:54.283 | ++ no_proxy=,,
derekhadarazs: we need to do something similar I'd say14:59
*** jpena|lunch is now known as jpena15:00
adarazsderekh: ack. let me see where we set that in tripleoci and what are the endpoints...15:00
jpenachem: I'm back, give me a sec to read the history15:02
derekhadarazs: no_proxy is set in overcloudrc15:03
derekhadarazs: export no_proxy=,,2001:db8:fd00:1000::1015:03
derekhadarazs: looks like it needs squar brackets around it15:04
adarazsderekh: no_proxy=,$MY_IP15:04
jpenachem EmilienM: we're already setting the env variable in t-h-t:
adarazsderekh: oh, I see.15:04
EmilienMjpena: right, what I showed to Sofer already15:04
EmilienMderekh, adarazs: FYI our Puppet OpenStack CI running IPv6 with Tempest is working: - we will need this patch in Tempest:
openstackgerritMichael Chapman proposed openstack/puppet-tripleo: Add keystone and db sync profiles
adarazsEmilienM: ack15:08
openstackgerritGiulio Fidente proposed openstack/tripleo-common: Create a test flavor for the pingtest VM
*** yamahata has joined #tripleo15:15
*** david-lyle has quit IRC15:15
*** david-lyle has joined #tripleo15:15
openstackgerritGiulio Fidente proposed openstack/tripleo-common: Use m1.small for ping test
*** liverpooler has quit IRC15:17
openstackgerritMerged openstack/tripleo-common: Use m1.small for ping test
openstackgerritGiulio Fidente proposed openstack/tripleo-common: Create a test flavor for the pingtest VM
adarazsderekh: what creates the overcloudrc?15:21
openstackgerritDougal Matthews proposed openstack/python-tripleoclient: Remove the demo flavour creation
adarazsderekh: I can't seem to find the part where we need to add the brackets15:21
derekhadarazs: probably python-tripleoclient, looking15:22
adarazsme too15:22
*** mbound has quit IRC15:22
adarazsderekh: okay, so the overcloud_ip needs a bracket if it's ipv615:24
derekhadarazs: yup, I believe so15:25
derekhadarazs: this seems to work no_proxy=,,[2001:db8:fd00:1000::110]15:25
adarazsderekh: you mean ::10 at the end.15:26
openstackgerritDmitry Tantsur proposed openstack/instack-undercloud: Remove unused scripts
derekhadarazs: yup, sorry, copy paste error15:26
adarazsderekh: wondering if somebody already wrote this "check if ipv6, bracket it if so" function in the python-tripleoclient. I'll look around.15:28
openstackgerritDmitry Tantsur proposed openstack/instack-undercloud: Remove unused elements
dtantsurrebased ^^15:28
openstackgerritRyan Brady proposed openstack/tripleo-common: Initial Integration of Mistral into TripleO
derekhadarazs: ok15:29
*** acruz has joined #tripleo15:30
openstackgerritDmitry Tantsur proposed openstack/instack-undercloud: Remove unused elements
*** jprovazn has joined #tripleo15:31
openstackgerritBen Nemec proposed openstack/tripleo-common: Remove ping test template from tripleo-common
openstackgerritJiri Stransky proposed openstack/tripleo-heat-templates: endpoint_map: bump nova api from 2.0 to 2.1
*** jprovazn has quit IRC15:33
*** jprovazn has joined #tripleo15:33
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Use m1.small for ping test
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Boot from volume in ping test
adarazsderekh: so keystone endpoint-list gives 503 now, while glance image-list works :)15:34
*** david-lyle_ has joined #tripleo15:34
*** david-lyle has quit IRC15:34
derekhadarazs: keystone endpoint-list could be using once of the internel IP's15:36
*** jaosorior has quit IRC15:36
adarazsderekh: so we need to add even more IPs to the no_proxy list?15:37
derekhadarazs: is that right ? should it in an IPv6 IP address ?15:37
derekhadarazs: ya, its failing to connect to as its not in no_proxy15:39
adarazsI see. but I'm not sure if that's correct or not. let me check an OSPd7 IPv6 deployment and compare :)15:39
dtantsurcould someone please merge ? it blocks a lot of documentation work..15:40
*** bvandenh has joined #tripleo15:40
dtantsurdocs only, 1x +2 already15:40
*** jroll has quit IRC15:41
*** jroll has joined #tripleo15:41
*** jroll has quit IRC15:41
*** jroll has joined #tripleo15:41
adarazsderekh: seems like the only endpoint that's not ipv6 is the keystone admin_url.15:42
adarazsnot sure why15:42
derekhadarazs: me neither15:42
adarazsderekh: I didn't alter anything from the default THT templates, so I'll look into what is not correct there.15:43
derekhadarazs: ok15:44
openstackgerritGiulio Fidente proposed openstack/python-tripleoclient: Generate a password for Redis and pass it as deployment parameter
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Ensure access to Redis is password protected
*** xinwu has joined #tripleo15:50
bnemecgfidente: Any particular reason you didn't merge ?15:50
gfidentebnemec, hey EmilienM is checking with kashyap15:51
EmilienMyeah but got no review15:51
gfidentebnemec, also, so the pingtest template moved to tripleo-ci?15:51
EmilienMif we're sure, let's go ahead15:51
bnemecgfidente: Yeah, but the removal patch missed the template so it didn't merge conflict existing patches, unfortunately.15:51
bnemecI pushed a removal patch to rectify that.15:51
gfidenteokay where is that so I +2 ? :)15:52
gfidenteand did your change for m1.small land in tripleo-ci?15:52
bnemecgfidente: Not yet.  That's
gfidentevoted there too15:53
bnemecThis move has been the bane of my existence the past couple of days. :-)15:53
bnemecYou don't realize how many CI patches you have open until they _all_ break at once.15:54
*** bvandenh has quit IRC15:54
gfidenteyeah like :)15:54
bnemecgfidente: It gave me an excuse to flex my -2 muscles. :-)15:55
adarazsderekh: correction: even in OSPd7 we have the keystone admin_url as ipv4. probably there's a good reason for that. but it seems we should add it to the no_proxy list as well then along with the bracketed IPv6 address if we want to make sure that keystone commands work as well.15:55
gfidentebnemec, except if I knew I'd have -2 your too instead of merging it15:56
bnemecgfidente: Yeah, that actually would have been good.  It didn't even occur to me that it was wrong until I saw that it had merged.15:56
gfidenteit was me off for a couple of days actually so I missed the stuff moved into the other repo15:57
*** tosky has quit IRC15:57
openstackgerritRyan Brady proposed openstack/tripleo-common: Add Heat Capabilities Actions
gfidentewait I have my chance for a -2 too15:58
*** tosky has joined #tripleo16:00
derekhadarazs: yup, makes sense16:00
openstackgerritGiulio Fidente proposed openstack-infra/tripleo-ci: Create a test flavor for the pingtest VM
openstackgerritMerged openstack/tripleo-docs: Don't repeat titles from the documents. Sphinx will add these.
openstackgerritBen Nemec proposed openstack/tripleo-common: Remove ping test template from tripleo-common
*** dshulyak has quit IRC16:08
openstackgerritAttila Darazs proposed openstack-infra/tripleo-ci: WIP: Use IPv6 on the gate jobs
gfidenteslagle, we need to port to stable/liberty too right?16:10
*** chem has joined #tripleo16:11
*** xinwu has quit IRC16:15
*** xinwu has joined #tripleo16:16
*** xinwu has quit IRC16:16
slaglegfidente: we could, i dont know that we have to16:16
openstackgerritMerged openstack/tripleo-heat-templates: compute: use live_migration_tunnelled
*** david-lyle_ is now known as david-lyle16:20
openstackgerritMerged openstack/tripleo-heat-templates: compute: include VIR_MIGRATE_TUNNELLED when doing VM shared storage
*** rbrady has quit IRC16:21
*** shivrao has joined #tripleo16:24
*** ebalduf has joined #tripleo16:24
*** shivrao_ has joined #tripleo16:26
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Properly setup DNS for the all CI jobs
rhalliseyderekh, ^ should address your comment16:27
rhalliseyderekh, I think I'll get the job working today ^.^16:28
*** shivrao has quit IRC16:28
*** shivrao_ is now known as shivrao16:28
derekhrhallisey: great ;-)16:28
*** ramishra has quit IRC16:30
*** jdob has quit IRC16:31
*** ramishra has joined #tripleo16:32
*** xinwu has joined #tripleo16:34
*** trown is now known as trown|lunch16:36
*** shivrao has quit IRC16:38
*** links has quit IRC16:38
*** shivrao has joined #tripleo16:39
*** xinwu has quit IRC16:42
*** leanderthal is now known as leanderthal|afk16:43
thrashif anybody has some time to take a look at the upgrade patchset in tripleo-common that would be awesome... starts here:
*** oshvartz has quit IRC16:45
*** athomas has quit IRC16:47
*** shivrao has quit IRC16:51
shadowerdtantsur: so I'm fine with but since I'm being removed from core, I should not give it a +2 or +A (and it already has a +1 from me)16:53
*** athomas has joined #tripleo16:53
dtantsurgot it, thanks anyway!16:53
dtantsurnot a couple more weeks hunting for the 2nd reviewer :D16:53
shadowerhope it'll be sooner :-)16:54
dtantsurthe first version of this patch appeared in the beginning of January, IIRC :)16:54
shadoweraltough I've seen worse16:55
*** aufi has quit IRC16:55
openstackgerritDmitry Tantsur proposed openstack/instack-undercloud: Remove unused scripts
openstackgerritMiles Gould proposed openstack/tripleo-common: Return list of new nodes from register_all_nodes
openstackgerritAthlan-Guyot sofer proposed openstack/puppet-pacemaker: Basic beaker one node test.
*** ifarkas has quit IRC17:03
openstackgerritDmitry Tantsur proposed openstack/python-tripleoclient: Switch to using node utilities from tripleo-common
openstackgerritMerged openstack/puppet-tripleo: Allow enabling authentication on haproxy.stats
*** dshulyak has joined #tripleo17:09
EmilienMbnemec: do we need backport in liberty? ^17:10
slagleEmilienM: yes17:10
openstackgerritEmilien Macchi proposed openstack/puppet-tripleo: Allow enabling authentication on haproxy.stats
openstackgerritAthlan-Guyot sofer proposed openstack/puppet-pacemaker: Add a service provider.
openstackgerritDmitry Tantsur proposed openstack/tripleo-docs: Extend the root device selection documentation
*** rasca has quit IRC17:12
*** mikelk has quit IRC17:12
*** jistr has quit IRC17:12
*** rasca has joined #tripleo17:13
adarazsfolks, I have a question: when we do an ipv6 deployment, we end up with all ipv6 endpoints except keystone's admin_url, which i still IPv4 (not sure why). we add the public_url's ip address to the no_proxy variable in overcloudrc, should we add the KeystoneAdminVip to it too?17:15
adarazsderekh, dsneddon ^17:16
*** xinwu has joined #tripleo17:16
*** ohamada_ has quit IRC17:17
dsneddonadarazs, The keystone admin VIP is on the control plane, which is IPv4-only. The keystone admin VIP is also on the Public VIP using IPv6, it just may not appear in the Keystone Catalog.17:18
dsneddonadarazs, Is it a problem for you that the Keystone admin VIP is IPv4?17:19
adarazsdsneddon: hm, so the problem is that when you do "keystone endpoint-list" it does this:17:20
adarazsDEBUG:keystoneclient.auth.identity.v2:Making authentication request to http://[2001:db8:fd00:1000::10]:5000/v2.0/tokens17:20
adarazsINFO:requests.packages.urllib3.connectionpool:Starting new HTTP connection (1): 2001:db8:fd00:1000::1017:20
adarazsDEBUG:requests.packages.urllib3.connectionpool:"POST /v2.0/tokens HTTP/1.1" 200 108917:20
adarazsDEBUG:keystoneclient.session:REQ: curl -g -i -X GET -H "User-Agent: python-keystoneclient" -H "Accept: application/json" -H "X-Auth-Token: {SHA1}a6b6acd7aaa7a5ae5abbec66b42c4808ca4bc5a4"17:20
adarazsdsneddon: so it asks on the IPv4 address... which, if not in the no_proxy list, won't work.17:20
*** dtantsur is now known as dtantsur|afk17:20
dsneddonadarazs, That sounds like a bug, can you file a BZ?17:20
adarazsdsneddon: bug in keystone client?17:21
dsneddonadarazs, Good question. I think this would actually be against rhel-osp-director, since we need to configure the no_proxy list in Director.17:21
derekhadarazs: dsneddon its probably a bug in tripleoclient, we probably need to add both to IP's to no_proxy in overcloudrc17:22
adarazsdsneddon: yeah, I'm there and I can make a change for it, I was just wondering if the expected thing is what keystone client does and we should add it to the no_proxy list, or not :)17:22
adarazsor it's a problem that the keystone client switches to the admin_url to get the endpoints... /me not sure how this is supposed to work...17:23
adarazsanyway, I'll create a bug for it to be sure. :)17:25
*** trown|lunch is now known as trown17:25
bnemecYeah, this is an outstanding problem.  You can have the same problem with all ipv4 net-iso if you're using a proxy.17:26
bnemecI've been meaning to fix it, but I keep getting distracted.17:26
adarazsbnemec: so should I add the admin vip to the no_proxy list?17:26
bnemecadarazs: I think so.17:27
adarazsokay. :)17:27
adarazsyay, /me fixes bugs in tripleo :)17:27
adarazsit has come to this.17:27
rdopieraadarazs: now you are responsible for that file17:28
rdopieraadarazs: all bugs in it are on you17:28
adarazsrdopiera: it's not too long at least :D17:28
rdopieranever fix bugs, they will think you know what you are doing17:29
derekhrdopiera: its not the file, its the whole repo he is responsible for17:30
rdopieramakes sense17:30
*** paramite has quit IRC17:32
* adarazs is not sure what he's getting himself into. :)17:32
*** lucasagomes is now known as lucas-afk17:33
*** miles is now known as mgould17:34
*** jdob_lt has joined #tripleo17:36
*** panda has quit IRC17:40
*** paramite has joined #tripleo17:40
*** panda has joined #tripleo17:41
ayoungwhat does openstack overcloud deploy --templates  (with no additional values) do?17:48
ayoungI   --templates [TEMPLATES]17:48
ayoung                        The directory containing the Heat templates to deploy17:48
ayoungbut both quickstark and both seem to pass it with no params17:49
*** derekh has quit IRC17:50
ayoungadarazs, keystoneclient.session:REQ: curl -g -i -X GET  is based on the AUTH_URL environment variable used to get the token in the first place.  From there on, it is using the service catalog17:51
ayoungadarazs, look in your overcloudrc file.  Comment out the existing AUTH_URL and make it the IPv6 one, and it should work.17:52
*** athomas has quit IRC17:52
adarazsayoung: it is the ipv6 one17:52
adarazs$ echo $OS_AUTH_URL17:52
ayoungadarazs, with v2?17:52
*** rcernin has quit IRC17:53
ayoungadarazs, and the endpoint in the SC was IPv6?17:53
adarazsayoung: it still started using the admin_url from the catalog.17:53
ayoungadarazs, oH...i MISREAD that17:53
ayoungadarazs, and what was the argument for putting the IPv4 address in there at all??17:54
ayoungadarazs, the IPv6 address should be the admin_url, not the ip_v417:56
adarazsayoung: we keep the control plane IPv4 and the keystone admin vip is on the control plane. I don't know why either decision was made in depth. something about IPMI not being ready for IPv6 i think.17:57
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Properly setup DNS for the all CI jobs
ayoungadarazs, OK, so, yeah, Keystone considers almost all operations as admin operations, and puts them on the Admin interface.  With V3 (which is what we should be using) it should be a unified URL17:59
ayoungbut that does not really solve your problem.17:59
openstackgerritAttila Darazs proposed openstack/python-tripleoclient: Add bracket around v6 overcloud IP in no_proxy
ayoungWe should be using DNS for this, not IP addresses17:59
adarazsayoung: yup. but for now we're fighting with brackets :D18:00
*** shivrao has joined #tripleo18:00
ayoungadarazs, I understand, and I lovethat we are getting IPv6 support.18:00
openstackgerritDan Prince proposed openstack/tripleo-puppet-elements: Add barbican-api-package to the overcloud.
openstackgerritBen Nemec proposed openstack/python-tripleoclient: Add password for HAProxy stats
*** olap has quit IRC18:13
*** akuznetsov has joined #tripleo18:16
*** tesseract- has quit IRC18:19
*** mgould has quit IRC18:20
openstackgerritPradeep Kilambi proposed openstack/tripleo-heat-templates: Move ceilometer api to run under wsgi apache
*** Marga_ has quit IRC18:40
*** jpena is now known as jpena|off18:41
*** electrofelix has quit IRC18:49
*** akuznetsov has quit IRC18:53
*** ebalduf has quit IRC18:53
openstackgerritJames Slagle proposed openstack/python-tripleoclient: Generate a password for Redis and pass it as deployment parameter
*** tosky has quit IRC19:09
*** acruz has quit IRC19:10
*** ayoung has quit IRC19:10
dsneddonEmilienM, Do you think the IPv6 changes in OPM are packaged and ready to test again?19:15
colonwqEmilienM, for 289027 I am workiing on an other puppet-ceph patch it will need to depend on. That patch should remove the hated execs.19:15
EmilienMdsneddon: not sure jayg did a build today19:16
EmilienMcolonwq: great.19:16
*** ryansb has quit IRC19:18
*** ryansb has joined #tripleo19:21
*** ryansb has joined #tripleo19:21
jaygEmilienM: we talking about master or liberty?  and is this in rdo?19:25
EmilienMjayg: rdo, master19:25
jaygI was awaiting word all the patches you needed were in, so no19:25
jaygis everything there?  I can sync now if so19:25
*** akuznetsov has joined #tripleo19:31
*** yamahata has quit IRC19:34
*** rbrady has joined #tripleo19:35
*** Marga_ has joined #tripleo19:36
dsneddondprince, Any chance you can reveiew and today? I'm getting concerned about getting these downstream in time.19:36
openstackgerritJames Slagle proposed openstack/instack-undercloud: Raise the mysql file limit
slagleEmilienM: go ahead man, laugh at my puppet ^19:36
EmilienMI was waiting for that day19:37
EmilienMslagle: your code looks good, I'm checking in puppetlabs-mysql if we can't do it out of the box. I hate writting specific code in tripleo that could be in puppet modules19:38
*** jdob_lt has quit IRC19:39
EmilienMslagle: commented19:43
slagleEmilienM: can i use the -> operator directly in the value for notify?19:44
slagleor do i just need to add that as a new line?19:44
EmilienMslagle: it has to be a new line19:44
*** jdob_lt has joined #tripleo19:45
openstackgerritJames Slagle proposed openstack/instack-undercloud: Raise the mysql file limit
EmilienMslagle: please remind me to give you a Puppetlabs sticker in Austin.19:46
slaglei don't think they'll let me wear that to the devops session on bash scripts19:46
EmilienMslagle: damn I forgot one thing in the review19:46
*** ryansb has quit IRC19:47
EmilienMslagle: commented again, should be fine after that.19:47
openstackgerritMerged openstack-infra/tripleo-ci: Use m1.small for ping test
openstackgerritJames Slagle proposed openstack/instack-undercloud: Raise the mysql file limit
slagleEmilienM: thanks19:48
*** ryansb has joined #tripleo19:49
*** ryansb has joined #tripleo19:49
openstackgerritZane Bitter proposed openstack/tripleo-heat-templates: Don't have separate protocols/ports for Keystone v3
*** akuznetsov has quit IRC19:54
*** Goneri has quit IRC19:56
openstackgerritAttila Darazs proposed openstack/python-tripleoclient: Fix no_proxy list generation for overcloudrc
EmilienMslagle: can I push over?20:02
EmilienMsome lint issues20:02
slagleEmilienM: sure20:02
*** jcoufal has quit IRC20:03
openstackgerritEmilien Macchi proposed openstack/instack-undercloud: Raise the mysql file limit
*** ayoung has joined #tripleo20:11
colonwqEmilienM, I have submitted 293693 for review, ridicule and making better.20:13
openstackgerritAttila Darazs proposed openstack/python-tripleoclient: Fix no_proxy list generation for overcloudrc
openstackgerritMichele Baldessari proposed openstack/python-tripleoclient: Clarify error when a template is missing
*** penick has joined #tripleo20:18
*** paramite has quit IRC20:21
*** jprovazn has quit IRC20:23
openstackgerritMerged openstack/puppet-pacemaker: Basic beaker one node test.
*** penick has quit IRC20:29
*** penick has joined #tripleo20:33
*** MaxPC has quit IRC20:42
EmilienMdsneddon: opm should be updated in master20:43
EmilienMlooking at delorean:
*** jayg is now known as jayg|g0n320:46
openstackgerritAttila Darazs proposed openstack/python-tripleoclient: Fix no_proxy list generation for overcloudrc
*** ianw has joined #tripleo20:49
*** ccamacho has quit IRC20:53
openstackgerritAttila Darazs proposed openstack-infra/tripleo-ci: WIP: Use IPv6 on the gate jobs
*** jdob_lt has quit IRC20:56
*** Goneri has joined #tripleo21:01
*** penick has quit IRC21:05
*** lblanchard has quit IRC21:08
openstackgerritMerged openstack/tripleo-heat-templates: endpoint_map: bump nova api from 2.0 to 2.1
*** shardy has quit IRC21:28
*** rdopiera has quit IRC21:28
*** dtantsur|afk has quit IRC21:31
*** pradk has quit IRC21:32
*** pino|work_ has joined #tripleo21:32
*** pino|work has quit IRC21:33
*** pblaho has quit IRC21:33
*** dprince has quit IRC21:34
*** pblaho has joined #tripleo21:34
*** colonwq has quit IRC21:34
*** pino|work has joined #tripleo21:36
*** pino|work_ has quit IRC21:37
*** dtantsur has joined #tripleo21:37
*** trown is now known as trown|outtypewww21:39
*** panda has quit IRC21:40
*** panda has joined #tripleo21:41
*** rbrady has quit IRC21:45
*** absubram has quit IRC21:49
*** absubram has joined #tripleo21:50
*** shivrao has quit IRC21:53
*** dshulyak has quit IRC21:54
*** shivrao has joined #tripleo21:56
*** shivrao has quit IRC22:00
openstackgerritMerged openstack/python-tripleoclient: Generate a password for Redis and pass it as deployment parameter
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Manage keystone initialization directly in t-h-t manifests
*** shivrao has joined #tripleo22:13
openstackgerritRyan Hallisey proposed openstack/tripleo-heat-templates: Hack to fix container gate
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
*** acruz has joined #tripleo22:17
bnemecI think puppet-pacemaker just broke our ha job. :-(22:17
bnemecEmilienM: ^22:18
EmilienMbnemec: ?22:19
EmilienMit was gate-tripleo-ci-f22-ha SUCCESS.22:19
EmilienMso I don't see how HA job can be broken22:19
*** r-mibu has quit IRC22:20
bnemecEmilienM: Hmm.  I'm seeing Error: Duplicate declaration: User[hacluster] is already declared in file /var/lib/heat-config/heat-config-puppet/c0b26626-4e80-436e-9226-28a5fc8e78ca.pp:73; cannot redeclare at /etc/puppet/modules/pacemaker/manifests/corosync.pp:121 on node overcloud-controller-0.localdomain22:20
*** r-mibu has joined #tripleo22:20
*** yamahata has joined #tripleo22:20
bnemecMaybe there was a colliding change in heat-config-puppet?22:21
bnemecOr probably tht, I guess.22:21
EmilienMhow to you explain gate-tripleo-ci-f22-ha passed?22:21
EmilienMwe are gating the module on TripleO Job22:21
bnemecEmilienM: No idea.  The hacluster user in tht was added almost a year ago:
EmilienMbnemec: I need to go, please email me if you need a revert and I'll do it.22:23
bnemecEmilienM: Will do, thanks22:23
EmilienMbnemec: I'll come back online later22:23
EmilienMso worst case I revert it from my phone if you email me22:23
EmilienMbut I would like to know why CI job passed.22:24
* EmilienM afk22:24
*** chlong has joined #tripleo22:32
*** penick has joined #tripleo22:34
*** penick has quit IRC22:36
*** penick has joined #tripleo22:40
*** yamahata has quit IRC22:48
*** TSCHAK has joined #tripleo22:49
*** ryansb has quit IRC22:49
openstackgerritBen Nemec proposed openstack/puppet-pacemaker: Revert "Basic beaker one node test."
*** gfidente has quit IRC22:51
*** TSCHAK_ has quit IRC22:52
*** ryansb has joined #tripleo22:52
*** ryansb has joined #tripleo22:52
openstackgerritmathieu bultel proposed openstack/tripleo-common: Add ssh option host key checking to no for automation
*** yamahata has joined #tripleo22:57
*** ooolpbot has joined #tripleo23:10
*** ooolpbot has quit IRC23:10
openstackLaunchpad bug 1558304 in tripleo "hacluster user conflict between t-h-t and puppet-pacemaker" [Critical,Triaged]23:10
*** yamahata has quit IRC23:18
*** akrivoka has quit IRC23:41
*** chlong has quit IRC23:41
*** TSCHAK_ has joined #tripleo23:44
*** TSCHAK has quit IRC23:47
slagleEmilienM: it's stable/liberty that it broke23:47
slagleEmilienM: oh, nm, looks like it's on master23:48
openstackgerritMerged openstack/tripleo-heat-templates: Ensure access to Redis is password protected
*** acruz has quit IRC23:51
*** dmacpher has joined #tripleo23:51
slagleoh i see, bnemec tracked it down23:51
*** penick has quit IRC23:57

Generated by 2.14.0 by Marius Gedminas - find it at!