| Kiall | once we shift the HP public cloud towards Designates V2 API, and we we need to rewrite anyway, Rally is the obvious choice.. We don't have the benefit of old code etc | 00:00 |
|---|---|---|
| boris-42 | Kiall: oh JMeter oh party hard=) | 00:00 |
| boris-42 | Kiall: btw we are going to support stuff like benchmarks for specific version | 00:01 |
| boris-42 | Kiall: cause now it's a bit tricky =) | 00:01 |
| ekarlso | Kiall: this stinks :P | 00:02 |
| Kiall | ekarlso: you broke it, you fix it ;) | 00:03 |
| ekarlso | Kiall: when passing in an endpoint should we rely on it being http://192.168.1.1:9001/v1 then or ? | 00:03 |
| Kiall | ehh... it depends on how keystone+you wrote the code I guess?> | 00:03 |
| ekarlso | Kiall: well i'm changing it so that that bit is only done when using u + | 00:04 |
| ekarlso | u + | 00:04 |
| ekarlso | sigh, u + p | 00:04 |
| Kiall | ccccccdugckeeifdgddnvdlvnucgkncckijnnnvridfv | 00:08 |
| Kiall | -_- | 00:08 |
| boris-42 | Kiall: ^ cat?) | 00:15 |
| Kiall | boris-42: VPN token ;) 1 time password | 00:15 |
| eandersson | btw thanks for the help Kialla. :) | 00:16 |
| ekarlso | Kialla ? :p :D | 00:16 |
| ekarlso | sounds like a girl :) | 00:16 |
| eandersson | I'll try to make some time to contribute back to Designate once I am back in Europe. | 00:16 |
| eandersson | lol | 00:16 |
| eandersson | *Kiall | 00:16 |
| Kiall | :) | 00:20 |
| openstackgerrit | Boris Pavlovic proposed openstack/designate: Add rally job related files https://review.openstack.org/134400 | 00:24 |
| ekarlso | boris-42: boris-42 fixed tghe stuff btw | 00:24 |
| boris-42 | Kiall: ekarlso ^ | 00:24 |
| boris-42 | ekarlso: nice =) | 00:24 |
| ekarlso | boris-42: it's now running -,,- | 00:24 |
| boris-42 | ekarlso: fucking yeah* =) | 00:25 |
| ekarlso | boris-42: but for V2 we'll need to make rally use a session -,,- | 00:25 |
| ekarlso | since it doesn't allow u to pass anything else :P | 00:25 |
| boris-42 | ekarlso: lol | 00:25 |
| boris-42 | ekarlso: okay | 00:25 |
| boris-42 | ) | 00:25 |
| boris-42 | ekarlso: so I made both patches in infra and in designate to init job * | 00:26 |
| openstackgerrit | Endre Karlson proposed openstack/python-designateclient: Make token + endpoints work correctly https://review.openstack.org/134410 | 00:26 |
| ekarlso | boris-42: clients you created are cached no ? | 00:26 |
| ekarlso | ^ boris-42 | 00:26 |
| boris-42 | ekarlso: ya | 00:27 |
| boris-42 | ekarlso: during the run of one scenario iteration* | 00:27 |
| boris-42 | ekarlso: otherwise you will get buthurt =) | 00:27 |
| boris-42 | ekarlso: cause they are not thread safe=) | 00:27 |
| ekarlso | :P | 00:27 |
| ekarlso | now if Kiall could sign off on it too... | 00:28 |
| ekarlso | it would be uber | 00:28 |
| Kiall | ekarlso: does it actually work? ;) | 00:28 |
| ekarlso | just sucks none of the other guys are here :p | 00:28 |
| boris-42 | ekarlso: lol | 00:28 |
| ekarlso | Kiall: yeah, I tseted with rally & the designate clietn | 00:28 |
| boris-42 | ekarlso: it's 3:30 a.m.=) | 00:28 |
| boris-42 | ekarlso: so people usually sleep at the moment in Russia=) | 00:28 |
| ekarlso | designate --os-token <toen> --os-endpoint http://192.168.27.100:9001 --debug domain-list | 00:28 |
| ekarlso | works fine ! | 00:28 |
| boris-42 | ekarlso: tested on devstack* lol | 00:29 |
| ekarlso | rally too actually :D | 00:29 |
| boris-42 | ekarlso: =) | 00:29 |
| ekarlso | boris-42: it's just 00:30 at Kiall's place so :p | 00:29 |
| Kiall | i.e. bedtime | 00:30 |
| ekarlso | Kiall: pffft :p | 00:30 |
| Kiall | Cya! Goodnight ;) | 00:31 |
| boris-42 | Kiall: see you | 00:31 |
| boris-42 | ekarlso: hmmm | 01:01 |
| boris-42 | ekarlso: how ot enable designate in devstack ? | 01:02 |
| eandersson | Take a look at this documentation | 01:04 |
| eandersson | http://designate.readthedocs.org/en/latest/devstack.html | 01:04 |
| boris-42 | eandersson: danke shoön | 01:12 |
| eandersson | np :) | 01:14 |
| boris-42 | eandersson: am I right that in designate there is only one WSGI service? | 01:35 |
| boris-42 | eandersson: cmd/api.py | 01:35 |
| *** nosnos has joined #openstack-dns | 01:43 | |
| boris-42 | ekarlso: pingngng | 02:04 |
| *** amcrn has quit IRC | 02:08 | |
| eandersson | There should only be the one afaik | 02:28 |
| eandersson | We use Central/Api/Sink | 02:28 |
| boris-42 | eandersson: so one more question | 03:01 |
| boris-42 | eandersson: where are rpc managers? is there something like https://github.com/openstack/cinder/blob/master/cinder/service.py#L121 | 03:03 |
| *** nosnos has quit IRC | 03:30 | |
| *** nosnos has joined #openstack-dns | 03:30 | |
| *** nosnos has quit IRC | 03:35 | |
| *** richm has quit IRC | 03:42 | |
| *** GonZo2K has quit IRC | 03:50 | |
| *** ryanpetrello has joined #openstack-dns | 04:06 | |
| *** ryanpetrello has quit IRC | 04:13 | |
| *** ryanpetrello has joined #openstack-dns | 04:18 | |
| *** nosnos has joined #openstack-dns | 04:27 | |
| *** ryanpetrello has quit IRC | 04:28 | |
| *** rmoe has quit IRC | 06:48 | |
| *** k4n0 has joined #openstack-dns | 07:08 | |
| *** nosnos has quit IRC | 08:00 | |
| *** nosnos has joined #openstack-dns | 08:03 | |
| *** jordanP has joined #openstack-dns | 08:30 | |
| openstackgerrit | Endre Karlson proposed openstack/python-designateclient: Make token + endpoints work correctly https://review.openstack.org/134410 | 08:50 |
| *** GonZo2K has joined #openstack-dns | 10:05 | |
| *** GonZo2K has quit IRC | 10:07 | |
| openstackgerrit | Endre Karlson proposed openstack/designate: Support secondary zones https://review.openstack.org/133682 | 10:09 |
| *** nosnos has quit IRC | 10:43 | |
| *** rediskin has joined #openstack-dns | 10:43 | |
| *** nosnos has joined #openstack-dns | 10:51 | |
| *** jmcbride has joined #openstack-dns | 11:06 | |
| *** Madkiss has quit IRC | 11:21 | |
| *** jmcbride has quit IRC | 11:32 | |
| openstackgerrit | Endre Karlson proposed openstack/designate: Support secondary zones https://review.openstack.org/133682 | 12:05 |
| *** nosnos has quit IRC | 12:33 | |
| *** EricGonczer_ has joined #openstack-dns | 12:33 | |
| openstackgerrit | Kiall Mac Innes proposed openstack/designate: Add tox check for missing pip requirements https://review.openstack.org/134516 | 12:37 |
| *** EricGonczer_ has quit IRC | 12:43 | |
| *** cbaesema has quit IRC | 12:45 | |
| *** EricGonczer_ has joined #openstack-dns | 12:53 | |
| *** russellb is now known as rustlebee | 13:02 | |
| *** EricGonczer_ has quit IRC | 13:03 | |
| openstackgerrit | Kiall Mac Innes proposed openstack/designate: WIP: Retry on Database Deadlocks https://review.openstack.org/134524 | 13:06 |
| *** EricGonczer_ has joined #openstack-dns | 13:15 | |
| *** ryanpetrello has joined #openstack-dns | 13:20 | |
| *** ryanpetrello has quit IRC | 13:30 | |
| *** betsy has joined #openstack-dns | 13:31 | |
| *** ryanpetrello has joined #openstack-dns | 13:33 | |
| boris-42 | Kiall: hm around ? | 13:37 |
| Kiall | Yep | 13:37 |
| openstackgerrit | Merged openstack/designate: Updated from global requirements https://review.openstack.org/134382 | 13:39 |
| boris-42 | Kiall: so I was wokring yesterday on integration osprofiler to designate | 13:45 |
| boris-42 | Kiall: and have question regarding to RPC services.. | 13:45 |
| Kiall | Sure | 13:45 |
| boris-42 | Kiall: so if we take a look at cinder | 13:46 |
| boris-42 | Kiall: I need something like that https://github.com/openstack/cinder/blob/master/cinder/service.py#L121-L122 | 13:46 |
| boris-42 | Kiall: so we can put tracing for all RPC managers methods calls | 13:46 |
| *** jordanP has quit IRC | 13:47 | |
| Kiall | Does it have do be the cls, or is the instance OK? https://github.com/openstack/designate/blob/master/designate/service.py#L67 | 13:47 |
| *** jordanP has joined #openstack-dns | 13:48 | |
| Kiall | There's also the create method on that class, which is effectively a factory.. so it could help? | 13:48 |
| Kiall | (I'm not 100% confident everything makes use of that factory, but we can fix that if there are cases which done) | 13:49 |
| Kiall | which don't* | 13:49 |
| Kiall | Also.. I kinda figured you would hook into RPC via oslo.messaging? | 13:51 |
| Kiall | brb | 13:51 |
| boris-42 | Kiall: soo there are 2 differnet separated parts | 13:51 |
| boris-42 | Kiall: one is done in serialization/desserialization of request (via oslo.messaging stuff) | 13:52 |
| boris-42 | Kiall: so actually it just makes next thing if profiler is initialized in caler (we will pass to request context profling info) | 13:53 |
| boris-42 | Kiall: on callee side if we see this info we will init profiler with passed data | 13:53 |
| boris-42 | Kiall: this is part one | 13:53 |
| boris-42 | Kiall: part 2 is that I would like to trace all calls of RPC service API | 13:54 |
| boris-42 | Kiall: and that is what I am asking about | 13:54 |
| boris-42 | Kiall: so endpoints is instance of what actually? | 13:55 |
| Kiall | effectively, the manager | 13:56 |
| Kiall | boris-42: L105 of https://github.com/openstack/designate/blob/master/designate/service.py should be what you need to wrap | 13:57 |
| Kiall | Humm | 13:57 |
| Kiall | Maybe | 13:57 |
| Kiall | oslo.messaging supports having multiple "endpoints" (historically, called managers in nova/cinder) | 13:58 |
| Kiall | here's a example of where the endpoint != the service class - https://github.com/openstack/designate/blob/master/designate/mdns/service.py#L37 | 14:05 |
| Kiall | that's currently the only example, and it's only using 1 endpoint but will expand.. | 14:05 |
| boris-42 | Kiall: hm hm | 14:06 |
| boris-42 | Kiall: so we can actually put on all classes @profiler.trace_cls() | 14:06 |
| boris-42 | Kiall: but it's not enough nice | 14:06 |
| boris-42 | Kiall: so what in service I should wrap? | 14:07 |
| boris-42 | Kiall: actually I need to wrap actually class (not base class) | 14:07 |
| Kiall | Ehh, I'm honestly not 100% sure | 14:08 |
| Kiall | well - for everything BAR designate-mdns, L105 of https://github.com/openstack/designate/blob/master/designate/service.py has a reference to the endpoint (aka "manager") cls before it's instantiated.. | 14:08 |
| boris-42 | Kiall: seems like I need to sleep more | 14:10 |
| boris-42 | Kiall: yep that is what I was actually looking for | 14:10 |
| Kiall | Yea, But, it work work (I think) for the designate-mdns service - or any service that makes use of multiple oslo.messaging managers? | 14:11 |
| *** nkinder has quit IRC | 14:11 | |
| boris-42 | Kiall: yep | 14:21 |
| boris-42 | Kiall: it will | 14:21 |
| boris-42 | Kiall: in any case you'll be able to play | 14:22 |
| boris-42 | Kiall: and see how it works | 14:22 |
| Kiall | are you sure? A quick look at the os-profiler code makes me think it wont :) But, yea.. It'll work for most cases and we can figure it out from there. | 14:22 |
| *** richm has joined #openstack-dns | 14:22 | |
| boris-42 | Kiall: why it won't lol? | 14:23 |
| boris-42 | Kiall: if it is isntance of class and it is instance of class it will | 14:23 |
| boris-42 | Kiall: in cinder and nova it works well lol | 14:23 |
| Kiall | the trace_cls method iterates the class methods and wraps them, which if your using oslo.messaging's multiple endpoints, means you can't just wrap the service class.. that doesn't actually have any of the RPC methods on it | 14:24 |
| Kiall | e.g. for designate-msdns, cls there will be the Service class here: https://github.com/openstack/designate/blob/master/designate/mdns/service.py while this is the "endpoint" https://github.com/openstack/designate/blob/master/designate/mdns/notify.py | 14:25 |
| Kiall | and - there can more more than 1 endpoint.. we don't use that (yet), but will | 14:25 |
| boris-42 | Kiall: heh | 14:26 |
| Kiall | hence why I expected os-profiler to hook straight into oslo-messaging, specifically the RPCDispatcher class | 14:26 |
| Kiall | that's where all incoming RPC calls "begin" | 14:26 |
| boris-42 | Kiall: i know but it's nicer to wrap methods that are exactly called | 14:27 |
| Kiall | Sure, but there could be 10s of different classes making up 1 RPC service :) | 14:27 |
| boris-42 | Kiall: huh maybe wrap all of them | 14:28 |
| boris-42 | Kiall: like iterate through multple endpoints and wrap them all | 14:28 |
| Kiall | Yep, but I'm not sure we have a way to do that right now :( | 14:28 |
| boris-42 | Kiall: or we can just add everywhere | 14:29 |
| boris-42 | @trace_cls | 14:29 |
| boris-42 | ^ | 14:29 |
| Kiall | That works I guess :) | 14:29 |
| boris-42 | okay let me publish something=) | 14:30 |
| *** EricGonczer_ has quit IRC | 14:32 | |
| * Kiall is scared of seeing the results ;) | 14:33 | |
| boris-42 | Kiall: lol | 14:33 |
| boris-42 | =) | 14:33 |
| * boris-42 Kiall: some day I'll stop working on Rally & OSprofiler and start working on improving OpenStack | 14:34 | |
| *** EricGonczer_ has joined #openstack-dns | 14:35 | |
| Kiall | eandersson: still looking at your issue... turning into a little bit of a nightmare -_- | 14:43 |
| *** ryanpetrello has quit IRC | 14:49 | |
| *** richm has quit IRC | 14:57 | |
| *** nkinder has joined #openstack-dns | 14:58 | |
| *** richm has joined #openstack-dns | 15:06 | |
| *** ryanpetrello has joined #openstack-dns | 15:12 | |
| *** jmcbride has joined #openstack-dns | 15:15 | |
| *** ryanpetrello has quit IRC | 15:17 | |
| Kiall | eandersson: I've filed https://bugs.launchpad.net/designate/+bug/1392762 for one part of the issue your seeing. | 15:23 |
| uvirtbot | Launchpad bug 1392762 in designate "Concurrent requests can cause designate-central to lock up" [High,Triaged] | 15:23 |
| openstackgerrit | Merged openstack/python-designateclient: Make token + endpoints work correctly https://review.openstack.org/134410 | 15:25 |
| openstackgerrit | Merged openstack/designate: Add rally job related files https://review.openstack.org/134400 | 15:26 |
| openstackgerrit | Merged openstack/designate: Add tox check for missing pip requirements https://review.openstack.org/134516 | 15:26 |
| boris-42 | Kiall: btw you can use rally for testing/reprodcuing such bugs in gates | 15:40 |
| *** nkinder is now known as nkinder_afk | 16:01 | |
| *** mwagner_lap has quit IRC | 16:05 | |
| *** paul_glass has joined #openstack-dns | 16:32 | |
| *** f1ller is now known as filler | 16:33 | |
| *** timsim has joined #openstack-dns | 17:03 | |
| *** mwagner_lap has joined #openstack-dns | 17:12 | |
| *** ryanpetrello has joined #openstack-dns | 17:15 | |
| eandersson | Thanks Kiall. | 17:29 |
| eandersson | I am feeling confident that we can mitigate the issue by simply adding more workers and nodes into the mix. | 17:39 |
| eandersson | For now anyway :p | 17:49 |
| *** ryanpetrello has quit IRC | 17:59 | |
| *** paul_glass has quit IRC | 18:04 | |
| *** ryanpetrello has joined #openstack-dns | 18:06 | |
| *** amcrn has joined #openstack-dns | 18:40 | |
| *** jordanP has quit IRC | 18:41 | |
| timsim | Kiall: Around? | 18:46 |
| *** ryanpetrello has quit IRC | 18:47 | |
| *** ryanpetrello has joined #openstack-dns | 18:52 | |
| *** paul_glass has joined #openstack-dns | 18:54 | |
| *** paul_glass has quit IRC | 18:55 | |
| ekarlso | Kiall: he's out | 19:01 |
| ekarlso | what's up ? | 19:01 |
| timsim | I think I may have figured it out...just a minute | 19:02 |
| *** ryanpetrello has quit IRC | 19:17 | |
| *** ryanpetrello has joined #openstack-dns | 19:22 | |
| *** MasterPiece has joined #openstack-dns | 19:37 | |
| *** MasterPiece has quit IRC | 19:58 | |
| openstackgerrit | Tim Simmons proposed openstack/designate: Add a reverse name column to the domains table https://review.openstack.org/134387 | 20:18 |
| *** ryanpetrello has quit IRC | 20:18 | |
| openstackgerrit | Tim Simmons proposed openstack/designate: Add a reverse name column to the domains table https://review.openstack.org/134387 | 20:21 |
| *** harmw_ is now known as harmw | 20:24 | |
| *** jmcbride has quit IRC | 20:46 | |
| *** ryanpetrello has joined #openstack-dns | 21:00 | |
| *** rustlebee is now known as russellb | 21:03 | |
| *** nkinder_afk is now known as nkinder | 21:08 | |
| openstackgerrit | Betsy Luzader proposed openstack/designate: Can't Access Default Pool through API https://review.openstack.org/134653 | 21:14 |
| *** paul_glass has joined #openstack-dns | 21:15 | |
| *** jmcbride has joined #openstack-dns | 21:19 | |
| *** ryanpetrello has quit IRC | 21:22 | |
| *** ryanpetrello has joined #openstack-dns | 21:25 | |
| betsy | Hmm - notice we have a new jenkins gate | 21:42 |
| betsy | non-voting, but mine failed it | 21:42 |
| *** ryanpetrello has quit IRC | 21:42 | |
| timsim | From the log, it doesn't appear to me to have anything to do with your patch: http://logs.openstack.org/53/134653/1/check/gate-rally-dsvm-designate-designate/ed5074a/console.html | 21:44 |
| openstackgerrit | Tim Simmons proposed openstack/designate: Add a reverse name columns to domains/recordsets https://review.openstack.org/134387 | 21:57 |
| ekarlso | any volunteers to test v2 cli ? | 21:57 |
| timsim | Maybe next week /yawns | 21:58 |
| *** EricGonczer_ has quit IRC | 22:03 | |
| *** timsim has quit IRC | 22:05 | |
| betsy | timsim: Yeah. I saw the log. I was just wondering when and why it got added | 22:20 |
| *** mwagner_lap has quit IRC | 22:25 | |
| *** EricGonczer_ has joined #openstack-dns | 22:25 | |
| *** paul_glass has quit IRC | 22:32 | |
| *** EricGonczer_ has quit IRC | 22:32 | |
| *** EricGonczer_ has joined #openstack-dns | 22:34 | |
| openstackgerrit | Boris Pavlovic proposed openstack/designate: Fix rally gates https://review.openstack.org/134662 | 22:34 |
| *** EricGonczer_ has quit IRC | 22:59 | |
| *** jmcbride has quit IRC | 23:03 | |
| boris-42 | GUYS you are so fast on review=) | 23:05 |
| boris-42 | https://review.openstack.org/#/c/134662/1 but it was not finished yet* | 23:05 |
| Kiall | lol | 23:06 |
| openstackgerrit | Boris Pavlovic proposed openstack/designate: Fix rally gates https://review.openstack.org/134662 | 23:07 |
| Kiall | betsy: that new job runs rally as part of the gate ;) https://github.com/stackforge/rally/ | 23:07 |
| Kiall | boris-42: finished? ;) | 23:08 |
| boris-42 | Kiall: I hope so | 23:08 |
| boris-42 | Kiall: actually benchmarks are going to fail unit ekarlso fix that issue with client* | 23:08 |
| boris-42 | Kiall: so there will be one more patch after fix from ekarlso that adds SLA checks* | 23:08 |
| boris-42 | Kiall: but for now we can merge this one | 23:08 |
| ekarlso | what issue now boris-42 ? | 23:08 |
| ekarlso | I thought I fixed it ?! | 23:08 |
| ekarlso | :P | 23:09 |
| boris-42 | ekarlso: oh | 23:09 |
| boris-42 | ekarlso: let see | 23:09 |
| Kiall | I think he did? https://github.com/openstack/python-designateclient/commit/cdf87e7fcaf2fe444904e785e1588bdb057f97de | 23:09 |
| boris-42 | ekarlso: ^_^ | 23:09 |
| boris-42 | ekarlso: Kiall I didn't know that!=) | 23:09 |
| Kiall | timsim: I spent the last 30 minutes trying to find how to make SQL Alchemy issue the REVERSE() function without doing it in code, and I'm totally shocked it's not supported -_- | 23:10 |
| boris-42 | Kiall: lol it happens=) | 23:12 |
| boris-42 | Kiall: btw it's faster to ask author =) | 23:12 |
| boris-42 | Kiall: zzzeek is his nickname+) | 23:12 |
| Kiall | Yea, I've talked to him a few times.. But I was on my phone, no IRC ;) | 23:13 |
| boris-42 | Kiall: heh=) | 23:13 |
| openstackgerrit | Endre Karlson proposed openstack/designate: MDNS part of Secondary zones https://review.openstack.org/133683 | 23:13 |
| boris-42 | ekarlso: Error <class 'designateclient.exceptions.Unknown'>: None uh OH | 23:25 |
| ekarlso | boris-42: where ? | 23:27 |
| boris-42 | ekarlso: your patch in rally that adds SLAs | 23:27 |
| boris-42 | ekarlso: i re-run check experimental | 23:27 |
| eandersson | Hmm, odd error. | 23:28 |
| eandersson | ResourceClosedError: This result object does not return rows. It has been closed automatically. | 23:28 |
| eandersson | Anyone seen this when trying to run multiple workers? | 23:29 |
| boris-42 | ekarlso: http://logs.openstack.org/39/133839/1/experimental/gate-rally-dsvm-designate-rally/b463f5a/rally-plot/sla.txt | 23:29 |
| eandersson | for Central | 23:29 |
| boris-42 | ekarlso: race lol? | 23:29 |
| ekarlso | boris-42: race ? :P | 23:30 |
| ekarlso | boris-42: I think the reason it is still bork is due to the fact that the fix is not in a "released" version yet of designateclient -,,- | 23:30 |
| ekarlso | oh, nvm that then | 23:30 |
| ekarlso | hmm | 23:30 |
| ekarlso | boris-42: how to get the full stack ? | 23:31 |
| boris-42 | ekarlso: https://github.com/boris-42/boris-42.github.io/blob/master/d.json#L122-L124 | 23:31 |
| Kiall | eandersson: I spotted that today while debugging your issue... It actually seems to be caused by the same issue :( | 23:31 |
| boris-42 | ekarlso: nope | 23:31 |
| boris-42 | ekarlso: cause dsvm is installing from master | 23:31 |
| boris-42 | ekarlso: at least it should lol | 23:31 |
| ekarlso | so it's deleting record. | 23:31 |
| boris-42 | ekarlso: seems like IT's race | 23:32 |
| boris-42 | ekarlso: part of iterations works part doesn't | 23:32 |
| boris-42 | ekarlso: it's not rally or client issue | 23:32 |
| boris-42 | Kiall: ^ so I think rally is good for such stuff=) | 23:32 |
| Kiall | eandersson: I have a fairly good idea on what the underlying issue is, it's just proving difficult to fix :( I filed a bug for it BTW - https://bugs.launchpad.net/designate/+bug/1392762 | 23:32 |
| uvirtbot | Launchpad bug 1392762 in designate "Concurrent requests can cause designate-central to lock up" [High,Triaged] | 23:32 |
| Kiall | boris-42: lol.. I'm sure it's the same damn bug -_- | 23:32 |
| ekarlso | :D | 23:32 |
| boris-42 | Kiall: so I can add SLA for all benchmarks and rally job will fail always.. | 23:33 |
| boris-42 | Kiall: or add everything except that benchmark | 23:33 |
| boris-42 | Kiall: on monday we will get a better HTML report (so you'll get inside it all traces) | 23:33 |
| boris-42 | Kiall: like this http://logs.openstack.org/44/131844/15/check/gate-rally-dsvm-rally/5001b1a/rally-plot/results.html.gz#/Dummy.dummy_exception_probability#failures | 23:34 |
| Kiall | Since it's non voting, I think it's OK to leave the broken run in.. We need to find + fix the issue anyway | 23:34 |
| boris-42 | Kiall: cause now you need to download full json and analyze it =( | 23:34 |
| *** EricGonczer_ has joined #openstack-dns | 23:34 | |
| boris-42 | Kiall: okay so I am going to add SLA check no failures | 23:34 |
| *** rediskin has left #openstack-dns | 23:35 | |
| openstackgerrit | Boris Pavlovic proposed openstack/designate: Fix rally gates https://review.openstack.org/134662 | 23:37 |
| Kiall | boris-42: that the winning patch? ;) | 23:37 |
| boris-42 | Kiall: ya it should be | 23:38 |
| boris-42 | =) | 23:38 |
| boris-42 | Kiall: I am usually doing for 40 patcheset's lol | 23:38 |
| boris-42 | =) | 23:38 |
| Kiall | lol... | 23:38 |
| boris-42 | Kiall: like this lol https://review.openstack.org/#/c/129060/ | 23:38 |
| boris-42 | =) | 23:38 |
| boris-42 | Kiall: not so fast as you guys=) | 23:38 |
| Kiall | ;) | 23:38 |
| Kiall | Okay.. Gone! It's late.. cyas! | 23:38 |
| ekarlso | Kiall: enjoy : ) | 23:39 |
| boris-42 | Kiall: =) | 23:39 |
| boris-42 | Kiall: see you | 23:39 |
| eandersson | Yep, read it last night Kiall. Thanks. | 23:39 |
| eandersson | I set up a few extra nodes to mitigate the issue. :p | 23:40 |
| *** nkinder has quit IRC | 23:40 | |
| boris-42 | ekarlso: hm so these warning about deleting resources http://logs.openstack.org/62/134662/2/check/gate-rally-dsvm-designate-designate/e08bcb5/console.html may be jsut cause of bug? | 23:41 |
| *** EricGonczer_ has quit IRC | 23:41 | |
| ekarlso | boris-42: which bug u mean ? | 23:43 |
| ekarlso | the deadlock one ? | 23:43 |
| boris-42 | ekarlso: idk =) | 23:43 |
| ekarlso | eandersson: u where from blizz no ? ^ | 23:43 |
| boris-42 | ekarlso: but for some reason part of iterations works=) | 23:43 |
| boris-42 | ekarlso: part not, as well during cleanup stuff | 23:43 |
| boris-42 | ekarlso: some resource can't be deleted cause max retries occur.. | 23:43 |
| ekarlso | boris-42: I think it's due to designate deadlocking | 23:43 |
| boris-42 | ekarlso: okay so I merged your patch in rally that adds SLAs | 23:44 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!