Monday, 2015-09-07

« Sunday, 2015-09-06 Index Tuesday, 2015-09-08 »
*** Qiming_ has joined #openstack-zaqar01:03
*** VeggieMeat_ has joined #openstack-zaqar02:08
*** dguerria has joined #openstack-zaqar02:13
*** kgriffsl has joined #openstack-zaqar02:13
*** dguerria is now known as dguerri02:13
*** dguerri has quit IRC02:13
*** dguerri has joined #openstack-zaqar02:13
*** kgriffsl is now known as kgriffs02:14
*** flwang1 has quit IRC02:15
*** flwang has quit IRC02:15
*** dguerri` has quit IRC02:15
*** kgriffs|afk has quit IRC02:15
*** VeggieMeat has quit IRC02:15
*** flaper87 has quit IRC02:15
*** flaper87 has joined #openstack-zaqar02:18
*** flwang has joined #openstack-zaqar02:24
*** flwang1 has joined #openstack-zaqar02:25
*** david-lyle has joined #openstack-zaqar03:06
*** Qiming__ has joined #openstack-zaqar03:07
*** Qiming_ has quit IRC03:07
*** vkmc has quit IRC03:08
*** david-ly_ has quit IRC03:08
*** vkmc has joined #openstack-zaqar03:14
*** Qiming__ has quit IRC03:44
openstackgerritFei Long Wang proposed openstack/zaqar: Fix wrong query for pools when creating queue  https://review.openstack.org/22084903:52
openstackgerritFei Long Wang proposed openstack/zaqar: Rename pool to pool_group for flavor  https://review.openstack.org/22087603:53
*** flwang1 has quit IRC05:04
*** Qiming__ has joined #openstack-zaqar05:23
*** diga has joined #openstack-zaqar05:44
*** flaper87 has quit IRC06:30
*** flaper87 has joined #openstack-zaqar06:30
*** Caleb-- has quit IRC08:00
*** Caleb-- has joined #openstack-zaqar08:12
*** flwang1 has joined #openstack-zaqar09:07
flwang1flaper87: morning09:09
flaper87flwang1: hello, sir!09:10
flaper87flwang1: hope you had a great weekend09:10
flaper87flwang1: if you're in the mood for reviews: https://review.openstack.org/#/c/218430/09:10
flwang1flaper87: yep, just delivered a session of zaqar at kiwi pycon09:10
flwang1flaper87: sure, i'm in the zaqar mode09:10
flaper87flwang1: AWESOME! How was it?09:12
flaper87Any feedback?09:12
flwang1not too much audience, TBH, just a question from Steve Baker09:13
openstackgerritMerged openstack/zaqar: Fix accurate links in README.rst  https://review.openstack.org/22058809:16
openstackgerritMerged openstack/zaqar: Refactoring to make Websocket driver available in v2  https://review.openstack.org/21843009:25
flwang1flaper87: just addressed your question in https://review.openstack.org/#/c/220849/09:26
flaper87flwang1: awesome, I'll get to it soon09:29
* flaper87 is trying to understand why Zaqar's gate doesn't log anything09:29
flwang1flaper87: for which gate? the rally one?09:30
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fix logging in Zaqar's gate  https://review.openstack.org/22092809:30
flaper87flwang1: all gates. If you look at the screen-zaqar logs, there's nothing there09:31
flaper87But Zaqar's config file is in debug mode09:31
flwang1flaper87: weird09:31
flwang1flaper87: btw, i'm review the code of flavor and i'm reading this line https://github.com/openstack/zaqar/blob/master/zaqar/storage/pooling.py#L18109:32
flwang1flavor = metadata and metadata.get('_flavor', None)09:32
flwang1so may i know we use '_flavor' as the key for flavor in metadata? do we have any naming rule for the reserved key in metadata?09:33
flaper87flwang1: kinda, the idea was to mark that key as "private" but I guess we could change that09:34
flwang1i'm going to add a 'token' key for email's subscription options, so that no spam for the unconfirmed subscriber09:34
flaper87flwang1: the main thing to keep in mind is that: Once the falvor is set, it cannot be changed09:34
flaper87to change it, a new queue needs to be created09:34
flaper87flavor migration will be implemented later09:34
flwang1flaper87: if so, i think we need to document this naming rule in our developer document and end user doc09:37
flwang1i'm good with _key :)09:38
flwang1so i'm going to add a _token key in email subscription's 'options', which will contains {'expired_time': '2015-09-08 12:34:67', 'confirmed': False}09:39
flwang1and we may need a small hacking for the endpoint GET /subscriptions/<sub_id> to support GET /subscriptions/<sub_id>?token=<UUID> to confirm the email subscription09:42
flwang1flaper87: does that make sense?09:42
flaper87flwang1: mmh, could you elaborate on what you need the token for?09:44
flaper87How's the token going to be generated?09:44
flwang1the token will be UUID generated by uuid lib09:45
flwang1after subscribed, the email address will receive an email with a link09:46
flaper87flwang1: ahh, nvm. I understand what you want to do now09:46
flaper87flwang1: should we do that in Mitaka?09:46
flwang1after click the link, the subscription will be activated/enabled from then09:46
flwang1flaper87: yep, sure.09:47
flaper87flwang1: I think it's useful but I think we should reduce the feature to the minimum in Liberty09:47
flaper87we can expand it in Mitaka09:47
flwang1flaper87: no problem09:47
flaper87That being said, I think it's a good thing to have09:47
flaper87flwang1: likely spec worthy09:47
flwang1flaper87: ok, i will open a blueprint to track it09:48
openstackgerritThomas Herve proposed openstack/zaqar: Notifications over websocket  https://review.openstack.org/20608010:01
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fix logging in Zaqar's gate  https://review.openstack.org/22092810:03
therveflwang1, I'm a bit dubious on the fact that this confirmation API ought to be in zaqar10:03
flaper87therve: I kinda think it's a nice, optional, thing to have10:04
flaper87but I'd assume that if I've authenticated to keystone and subscribed an email, I should be trusted as a source thus the confirmation shouldn't be needed10:05
therveflaper87, I don't know, not sure going into an email service is the best way forward10:05
therveIt also crosses abstractions barriers, which always suspicious10:06
therve(subscriptions are generic, but confirmation are pretty much email specific)10:06
flwang1therve: i can see your point10:06
flwang1but my concern is from a cloud provider view10:06
flaper87flwang1: https://bugs.launchpad.net/zaqar/+bug/1489879 didn't you have a patch for that?10:07
openstackLaunchpad bug 1489879 in zaqar "Got 500 error when addding new mongodb pool" [Low,Triaged] - Assigned to Fei Long Wang (flwang)10:07
flwang1as a public cloud provider, the tenant may use this service to send spam10:07
therveI'm not arguing the usefulness of the feature10:07
therveI'm arguing against  it being in zaqar10:08
flwang1flaper87: not yet, but i can do it soon :)10:09
flwang1therve: i see, but IIRC, SNS has the same feature10:09
flaper87flwang1: no need, I just seemed to remember there was one10:10
flaper87flwang1: yes, it does.10:10
flwang1flaper87: got10:10
flaper87therve: where would you have it instead?10:10
therveflwang, SNS has a dedicated confirm API though, not a hack on GET subscriptions10:11
therveflaper87, Another service10:11
therveI feel that even the webhook implementation ought to be separate10:11
therveZaqar's one is very fragile and needs lots of work10:12
flwang1therve: oh, hacking GET just my init thought, we can define a good one for sure10:12
therveLike, doesn't mistral allows sending email?10:12
flaper87therve: but wait, webhook and emails are just plugins for the notification API which is part of Zaqar10:15
flaper87not sure why you think those don't belong in Zaqar10:15
therveMostly because they are a full time job by themselves10:16
flaper87(brb, lunch, but we'll talk a bit more about this later today if you're around)10:18
therveSure thing10:19
*** diga has quit IRC10:20
*** Qiming__ has quit IRC10:53
*** flwang1 has quit IRC11:08
*** flwang1 has joined #openstack-zaqar11:11
*** flwang1 has quit IRC11:18
*** flwang1 has joined #openstack-zaqar11:22
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fix logging in Zaqar's gate  https://review.openstack.org/22092811:35
flwang1flaper87: back?11:37
flaper87flwang1: yes11:39
flwang1flaper87: a quick question, why are we keeping the tenant id in flavor?11:39
flwang1IMHO, flavor should be a global resource like pool11:40
flwang1that said, it doesn't belong any tenant technically11:40
flaper87flwang1: I was wondering the same the other day while I was implementing the sqlalchemy thing11:41
flaper87and I believe I had an answer for that11:41
flaper87I don't remember anymore11:41
flaper87:P11:41
flwang1flaper87: ok, i believe there's an answer :D11:42
flaper87I believe the reason behind that is that an admin may want to create a flavor that is just visible by one tenant11:42
flaper87for billing porpuses11:43
flaper87but I'm just guessing here11:43
flaper87TBH11:43
flaper87I believe the queries for getting flavors don't even support that scenario11:43
flaper87(yet_11:43
flaper87(yet)11:43
flwang1flaper87: i can  see your point but not really clearly11:44
flwang1flaper87: so should we hide the project id as long as we open the flavors list11:44
flwang1since it may confuse the end user?11:44
flaper87we can hide it for now, yes.11:45
flaper87Don't remove it, though.11:45
flaper87and, pleeeeeeeease, file a bug11:45
flaper87:D11:45
flaper87ok, brb. Need to finish something11:45
* flaper87 mutes emails and irc11:45
*** flwang1 has quit IRC11:49
*** flwang1 has joined #openstack-zaqar12:00
*** Qiming_ has joined #openstack-zaqar12:12
*** exploreshaifali has joined #openstack-zaqar12:15
*** dynarro has joined #openstack-zaqar12:17
*** gzhai has joined #openstack-zaqar12:22
openstackgerritFlavio Percoco proposed openstack/zaqar: Let devstack redirect output  https://review.openstack.org/22092812:28
openstackgerritFlavio Percoco proposed openstack/zaqar: Let devstack redirect output  https://review.openstack.org/22092812:30
*** vkmc has quit IRC12:32
*** vkmc has joined #openstack-zaqar12:32
*** Qiming_ has quit IRC12:33
*** Qiming__ has joined #openstack-zaqar12:33
openstackgerritFei Long Wang proposed openstack/zaqar: Hide project id for flavor  https://review.openstack.org/22098012:38
therveflaper87, couple of last comments I hope on https://review.openstack.org/#/c/209910/12:40
*** exploreshaifali has quit IRC13:06
flaper87therve: thanks, will address those now13:43
flaper87therve: https://review.openstack.org/#/c/220928/13:43
openstackgerritFlavio Percoco proposed openstack/zaqar: Policy support  https://review.openstack.org/20991013:49
*** gzhai has quit IRC13:52
flaper87vkmc: btw, the v2 websocket port landed14:01
flaper87:D14:01
flaper87w000h000014:01
vkmcwohooo14:01
vkmc:D14:01
angvpflaper87: what about the pizza?14:14
dynarrovkmc, yay!!!14:15
vkmcdynarro, :D14:15
angvphai14:15
vkmcdynarro, are you joining us for the pizza marathon? flaper87 is hosting it14:15
dynarrovkmc, pizza? marathon? what?!14:16
vkmchaha14:17
vkmcangvp, I blame you for this idea14:19
vkmcangvp, so you tell her14:20
angvpvkmc: done14:20
dynarroOMG! that would be the best marathon ever14:25
dynarrovkmc, angvp why italians haven't thought about that?!14:26
dynarroI'll propose the idea haha14:26
angvpbecause i'd be a problem and north and south gonna fight again14:27
dynarrohaha14:27
angvpit will be*14:27
dynarroI could walk for more that 7 km if they give me pizza on the way...or cookies14:29
dynarroLOL14:30
vkmcangvp, do Italians also have faina?14:30
angvpvkmc: si, farinata si chiama14:30
vkmcangvp, I love Italians14:31
angvpvkmc: sorry, it's called "farinata" and in liguria are supposed to be theee place of the farinata14:31
vkmcthis week in Buenos Aires I plan to get my faina three-month supply14:32
angvpI just have been hours in liguria and I didn't knew the existence of farinata by that moment14:32
angvpvkmc: don't you have faina at bhi ?14:32
vkmcangvp, no :'(14:32
vkmcwe are a different kind of Argentinians down here14:32
dynarroseems that angvp is right14:35
dynarroI haven't hear of it before14:36
dynarroheard*14:36
vkmcits like an add-on for pizza14:36
vkmcits amazing14:36
dynarroneed to try that =)14:40
* flaper87 has no idea what these crazy people are talking about14:49
therveflaper87, Did you notice that zaqar functional tests doesn't use the devstack spawned version?14:49
therveI guess it's because of that trick in the setup14:49
flaper87therve: you mean the one we have in the code base?14:50
therveflaper87, the one what? :)14:50
flaper87therve: erm, sorry. I meant to say if you mean it's not using the plugin14:50
therveflaper87, It's using the devstack plugin, and it spawns it properly14:51
therveflaper87, But the tests spawn a in memory version of the server AFAIU14:51
flaper87therve: ah, yeah, the functional tests do. But those run in the py27 gate14:52
flaper87not the dsvm one14:52
therveHum14:52
flaper87The py27 gate doesn't have a running zaqar14:52
therveflaper87, Why isn't there anything the dvsm logs then, even with your patch?14:52
flaper87which is why the functional tests spawn a separate process14:53
flaper87therve: I'm still trying to find out that14:53
flaper87:P14:53
therveAh :)14:53
flaper87I think that patch is good to land, though.14:53
flaper87but there's more work to do14:53
therveYeah14:54
therveIt doesn't look like it's running any tests14:54
flaper87Courtesy ping: vkmc therve ryansb flwang dynarro jasondotstar15:00
therveMeeting?15:01
vkmcmeeting!15:01
*** diga has joined #openstack-zaqar15:03
vkmcdiga, if interested, we are having the weekly meeting15:07
vkmcin #openstack-meeting-315:07
digasure15:07
digalet me join15:07
vkmcdiga, sorry for the delay on the review of the pools patch for the client, it merged some weeks ago15:07
vkmcnow that we have a new core, reviews will be quicker :)15:07
digavkmc: that's great15:08
openstackgerritThomas Herve proposed openstack/zaqar: Enable tempest run in the gate  https://review.openstack.org/22104815:17
*** Qiming__ has quit IRC15:28
flaper87therve: so, now that I remember. I believe one of the reasons tempest was disabled is because it wasn't entirely compatible and because we wanted to use zaqarclient instead for functional tests15:28
flaper87That is, work on improving zaqarclient's functional tests and run them on the dsvm gate15:29
flaper87not sure how good/bad that idea is15:29
flaper87therve: also, you may want to make your patch depend on mine to get more logs15:29
therveflaper87, I think I did?15:31
therveI thought I did :)15:32
flaper87therve: :D15:32
therveYeah looks right, we'll see if it worked15:34
therveflaper87, I can see why we wouldn't want to use tempest. But let's not have useless jobs then15:36
flaper87therve: I fully agree, I think it just fell off of our radar15:37
flaper87lets see what your patch results are15:37
flaper87erm15:37
flaper87lets see what the results of your patch are15:37
* flaper87 can't english15:37
vkmcyour English is on holidays15:38
vkmchahah15:38
flaper87vkmc: LOOOOOL15:39
flaper87pretty much15:39
vkmcflaper87, I was going to reply that here https://twitter.com/flaper87/status/64081035267073228815:40
vkmcbut I decided to save it for later15:40
flaper87:P15:40
therveAhahah15:40
vkmchahaha15:40
vkmcI can't say anything though, my English lives on holidays15:41
flaper87I gave up with twitter15:41
flaper87it's impossible to not have typos there15:41
flaper87especially when one keeps ninja-editing tweets15:42
flaper87sometimes I can't even understand my own tweets15:42
flaper87dynarro: when do you think you'll be able to get these patches green? https://review.openstack.org/#/c/209238/15:43
flaper87do you need help with those?15:43
flaper87vkmc: https://review.openstack.org/#/c/202248/15:44
dynarroflaper87, I'm trying to figure out why that isn't working15:45
flaper87vkmc: can you look at this too? https://review.openstack.org/#/c/220928/15:46
flaper87therve: ^15:46
flaper87I'm kinda tempte to approve that w/ just 1 +215:46
flaper87because I need it to figure out what's going on in the client's gate15:46
openstackgerritDoraly Navarro proposed openstack/python-zaqarclient: Added v1 folder  https://review.openstack.org/21615915:46
flaper87I guess I can use Depends-On15:46
flaper87meh15:46
therveflaper87, What does --master do?15:47
flaper87therve: it just enables uwsig's built-in workers management15:48
therveHum ,ok15:49
flaper87It's not really necessary, I added it to suppress a warnning saying it was not running in master mode15:49
flaper87therve: https://uwsgi-docs.readthedocs.org/en/latest/Glossary.html?highlight=master15:49
therveAH ok15:50
therve+1 then15:50
flaper87therve: so, I think we'll need a couple of patches between project-config, zaqar and zaqarclient to make this work properly15:59
* flaper87 tries something16:00
therveYeah sounds like it16:00
therveSo tempest tests did run and failed16:00
flaper87therve: I think I have an idea of how to make this work.16:00
therveThe question is whether we want to fix them or simply remove the builds16:00
flaper87I think we could have a single job template in project-config that works well for zaqarclient functional tests and tempest16:01
flaper87then have the required things installed by gate_hook16:01
therveDo we care about tempest though? Having tests in-tree is much easier16:01
flaper87therve: not much, in fact, I'd like to convert those gates to zaqarclient as part of the above idea16:01
flaper87but still make it super easy to add tempest back if the tests are updated16:02
flaper87lemme take this one16:02
therveOK16:02
flaper87therve: if the above doesn't work, I'll bounce it back to you16:03
flaper87ahahahha16:03
therve:D16:03
openstackgerritFlavio Percoco proposed openstack/zaqar: Let devstack redirect output  https://review.openstack.org/22092816:17
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106816:18
*** dynarro has quit IRC16:20
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106816:23
openstackgerritFlavio Percoco proposed openstack/python-zaqarclient: WIP: fixing gate  https://review.openstack.org/22107016:24
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106816:47
openstackgerritFlavio Percoco proposed openstack/python-zaqarclient: WIP: fixing gate  https://review.openstack.org/22107016:48
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106816:51
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106816:57
*** stanchan has joined #openstack-zaqar17:11
*** csoukup has joined #openstack-zaqar17:13
*** diga has quit IRC17:22
*** boris-42 has quit IRC17:30
*** dguerri is now known as dguerri`17:50
*** dynarro has joined #openstack-zaqar17:56
*** davideagnello has joined #openstack-zaqar18:08
openstackgerritDoraly Navarro proposed openstack/python-zaqarclient: Added support to claims  https://review.openstack.org/21316518:09
openstackgerritDoraly Navarro proposed openstack/python-zaqarclient: Added support to notifications  https://review.openstack.org/20923818:09
openstackgerritDoraly Navarro proposed openstack/python-zaqarclient: Added v1 folder  https://review.openstack.org/21615918:09
*** exploreshaifali has joined #openstack-zaqar18:46
*** dynarro has quit IRC18:47
*** davideagnello has quit IRC18:50
*** zxyf has joined #openstack-zaqar19:07
*** zxyf has left #openstack-zaqar19:13
*** zxyf has joined #openstack-zaqar19:16
*** zxyf has quit IRC19:22
*** zxyf has joined #openstack-zaqar19:23
*** exploreshaifali has quit IRC19:34
*** flwang1 has quit IRC19:34
*** flwang1 has joined #openstack-zaqar19:34
*** boris-42 has joined #openstack-zaqar19:54
*** flwang1 has quit IRC20:10
*** dynarro has joined #openstack-zaqar20:12
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106820:15
*** csoukup has quit IRC20:26
*** Isaac has joined #openstack-zaqar20:35
*** zxyf has quit IRC20:36
*** Isaac is now known as zxyf20:39
openstackgerritDoraly Navarro proposed openstack/python-zaqarclient: Added support to v2  https://review.openstack.org/20923820:41
*** flwang1 has joined #openstack-zaqar20:43
*** flwang1 has quit IRC20:47
*** davideagnello has joined #openstack-zaqar20:51
*** flwang1 has joined #openstack-zaqar20:51
*** dynarro has quit IRC20:53
*** davideagnello has quit IRC20:56
openstackgerritFlavio Percoco proposed openstack/zaqar: WIP: Fixing gate  https://review.openstack.org/22106821:14
*** david-lyle has quit IRC22:39
*** david-lyle has joined #openstack-zaqar22:44
*** gzhai has joined #openstack-zaqar22:48
*** davideagnello has joined #openstack-zaqar22:52
*** davideagnello has quit IRC22:57
*** davideagnello has joined #openstack-zaqar23:05
*** zxyf has quit IRC23:19
openstackgerritFei Long Wang proposed openstack/zaqar: Closes-Bug: #1489879  https://review.openstack.org/22112223:26
openstackbug 1489879 in zaqar "Got 500 error when addding new mongodb pool" [Low,Triaged] https://launchpad.net/bugs/1489879 - Assigned to Fei Long Wang (flwang)23:26
*** Qiming__ has joined #openstack-zaqar23:41
« Sunday, 2015-09-06 Index Tuesday, 2015-09-08 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!