Monday, 2018-01-08

« Sunday, 2018-01-07 Index Tuesday, 2018-01-09 »
*** slaweq has joined #openstack-lbaas00:11
*** slaweq has quit IRC00:15
*** sbalukoff_ has quit IRC00:15
*** sbalukoff_ has joined #openstack-lbaas00:28
*** atoth has quit IRC01:18
*** annp has joined #openstack-lbaas01:54
openstackgerritBar Elharar proposed openstack/python-octaviaclient master: Code cleanup  https://review.openstack.org/53004502:51
openstackgerritBar Elharar proposed openstack/python-octaviaclient master: Improve unit testing scalability  https://review.openstack.org/53056702:51
openstackgerritBar Elharar proposed openstack/python-octaviaclient master: Improve unit tests results with subTest  https://review.openstack.org/53125702:51
*** sanfern has quit IRC02:53
*** cyberspac69 has joined #openstack-lbaas03:08
*** cyberspac69 has quit IRC03:10
*** kong has quit IRC03:17
*** kong has joined #openstack-lbaas03:17
*** yamamoto has joined #openstack-lbaas03:58
*** sanfern has joined #openstack-lbaas03:59
*** gans has joined #openstack-lbaas04:02
*** yamamoto has quit IRC04:03
openstackgerritSanthosh Fernandes proposed openstack/octavia master: ACTIVE-ACTIVE: Initial distributor data model  https://review.openstack.org/52885004:27
openstackgerritSanthosh Fernandes proposed openstack/octavia master: [WIP] L3 ACTIVE-ACTIVE Data model impact  https://review.openstack.org/52472204:27
*** yamamoto has joined #openstack-lbaas05:04
*** yamamoto has quit IRC05:12
*** Alex_Staf has joined #openstack-lbaas06:35
*** reedip has quit IRC06:55
*** ltomasbo has joined #openstack-lbaas06:59
*** pcaruana has joined #openstack-lbaas07:04
*** reedip has joined #openstack-lbaas07:07
*** ctracey has quit IRC07:13
*** fyxim has quit IRC07:13
*** fyxim_ has joined #openstack-lbaas07:13
*** ctracey has joined #openstack-lbaas07:13
*** b_bezak has joined #openstack-lbaas08:07
*** tesseract has joined #openstack-lbaas08:09
*** nmagnezi has joined #openstack-lbaas08:45
openstackgerrithang cheng proposed openstack/octavia master: TrivialFix: fix a typo in comment  https://review.openstack.org/53169709:07
*** rcernin has quit IRC09:07
openstackgerrithang cheng proposed openstack/octavia master: TrivialFix: fix a typo in comment  https://review.openstack.org/53169709:09
*** gans has quit IRC09:16
*** sanfern has quit IRC09:18
*** sanfern has joined #openstack-lbaas09:26
*** nmagnezi_ has joined #openstack-lbaas09:28
*** gans has joined #openstack-lbaas09:30
*** nmagnezi has quit IRC09:30
*** nmagnezi_ is now known as nmagnezi09:30
*** gans has quit IRC09:32
*** salmankhan has joined #openstack-lbaas10:01
*** sanfern has quit IRC10:20
-openstackstatus- NOTICE: zuul has been restarted, all queues have been reset. please recheck your patches when appropriate10:25
*** annp has quit IRC10:32
*** numans has quit IRC11:06
*** numans has joined #openstack-lbaas11:08
nmagnezirm_work, o/11:13
openstackgerritMerged openstack/octavia-dashboard master: Make whitelist externals os agnostic  https://review.openstack.org/52852911:40
*** sanfern has joined #openstack-lbaas11:48
*** atoth has joined #openstack-lbaas12:24
*** sanfern has quit IRC12:26
*** rcernin has joined #openstack-lbaas13:00
*** rcernin has quit IRC13:16
*** pcaruana has quit IRC13:25
*** pcaruana has joined #openstack-lbaas13:25
*** yamamoto has joined #openstack-lbaas13:28
*** leitan has joined #openstack-lbaas13:36
*** salmankhan1 has joined #openstack-lbaas13:55
*** yamamoto has quit IRC13:56
*** yamamoto has joined #openstack-lbaas13:57
*** salmankhan has quit IRC13:59
*** salmankhan1 is now known as salmankhan13:59
*** yamamoto has quit IRC14:01
*** dayou has joined #openstack-lbaas14:03
*** dayou has quit IRC14:03
*** dayou has joined #openstack-lbaas14:03
openstackgerritSanthosh Fernandes proposed openstack/octavia master: [WIP] L3 ACTIVE-ACTIVE Data model impact  https://review.openstack.org/52472214:13
*** sanfern has joined #openstack-lbaas14:13
*** dayou has quit IRC14:28
*** yamamoto has joined #openstack-lbaas14:42
*** yamamoto has quit IRC14:47
*** AlexeyAbashkin has joined #openstack-lbaas14:53
*** dayou has joined #openstack-lbaas14:55
*** Alex_Staf has quit IRC15:01
*** leitan has quit IRC15:02
*** AlexeyAbashkin has quit IRC15:13
*** yamamoto has joined #openstack-lbaas15:27
*** b_bezak has quit IRC15:31
*** yamamoto has quit IRC15:32
*** leitan has joined #openstack-lbaas15:51
*** leitan_ has joined #openstack-lbaas15:52
*** leitan has quit IRC15:52
*** yamamoto has joined #openstack-lbaas15:53
*** yamamoto has quit IRC15:53
*** AlexeyAbashkin has joined #openstack-lbaas16:02
*** AlexeyAbashkin has quit IRC16:06
*** armax has joined #openstack-lbaas16:21
*** longstaff has joined #openstack-lbaas16:29
openstackgerritMerged openstack/octavia master: TrivialFix: fix a typo in comment  https://review.openstack.org/53169716:32
openstackgerritMerged openstack/octavia master: Use common functional tox jobs  https://review.openstack.org/52857816:48
openstackgerritMerged openstack/octavia master: Force SELinux context for amphora keepalived process  https://review.openstack.org/52707316:48
*** pcaruana has quit IRC16:48
*** yamamoto has joined #openstack-lbaas16:53
*** longstaf_ has joined #openstack-lbaas16:54
*** longstaff has quit IRC16:58
*** slaweq has joined #openstack-lbaas17:00
*** yamamoto has quit IRC17:02
*** longstaff has joined #openstack-lbaas17:04
*** longstaf_ has quit IRC17:07
*** slaweq has quit IRC17:11
*** longstaff has quit IRC17:24
sanfernhi johnsom17:25
johnsomHi sanfern17:25
*** longstaff has joined #openstack-lbaas17:25
*** longstaff has quit IRC17:27
sanfernComment on assigning VIP to loopback device and making this change to persistence, i need to use Jinja template, do I need to create one for interface file or ... ?17:27
*** longstaff has joined #openstack-lbaas17:27
johnsomThey are all here already: https://github.com/openstack/octavia/tree/master/octavia/amphorae/backends/agent/api_server/templates17:28
sanfernyes17:28
sanfernShould I use existing one17:29
johnsomSo, I would expect you would either update those, or create a new one if the interface doesn't match one we already have.17:29
sanfernplug_vip_ethX.conf.j217:29
johnsomI think I asked if loopback is right or if it should be a device type of "dummy", did we get an answer on that?17:30
sanfernwe tested creating new loopback device acted like dummy interface17:31
johnsomI think it's common for the BGP stuff to go on a device of type dummy instead of loopback17:31
sanfernyeah17:31
johnsomloopback is special and I'm a bit concerned about using it17:31
johnsomIt would make it harder I think as we just copy over the default loopback interface configuration file from the OS distro17:32
sanfernok17:33
sanfernWhat is any alternative solution u recommend use dummy interface17:34
sanfernDo we need to add this dummy interface in ExaBGP element ?17:34
johnsomYeah, I think if we just use a dummy device instead of lo it will be easier to develop.  You would use one of the existing jinja templates, or create a new one if needed, and instead of "lo" you would use "dummy#" where # is the interface number.17:36
johnsomShouldn't change the element if I remember correctly17:36
sanfernok let me try in lab17:37
johnsomOk17:37
sanfernI will create new interface template17:37
johnsomWorst case, maybe we have to add "dummy" to /etc/modules to get the kernel module, but I'm pretty sure it's already enabled17:37
sanfernok17:38
*** yamamoto has joined #openstack-lbaas17:43
*** yamamoto has quit IRC17:43
*** b_bezak has joined #openstack-lbaas17:52
*** AlexeyAbashkin has joined #openstack-lbaas17:54
*** sanfern has quit IRC17:56
*** b_bezak has quit IRC17:56
*** AlexeyAbashkin has quit IRC17:58
*** slaweq has joined #openstack-lbaas18:05
*** salmankhan has quit IRC18:16
*** slaweq has quit IRC18:17
*** slaweq has joined #openstack-lbaas18:21
*** slaweq has quit IRC18:31
*** slaweq has joined #openstack-lbaas18:34
*** tesseract has quit IRC18:35
*** mestery has quit IRC18:40
*** longstaff has quit IRC18:42
*** yamamoto has joined #openstack-lbaas18:44
*** mestery has joined #openstack-lbaas18:44
*** yamamoto has quit IRC18:52
*** slaweq has quit IRC19:01
*** yamamoto has joined #openstack-lbaas19:01
*** slaweq has joined #openstack-lbaas19:02
*** leitan has joined #openstack-lbaas19:05
*** leitan_ has quit IRC19:05
*** slaweq has quit IRC19:06
*** slaweq has joined #openstack-lbaas19:09
*** sanfern has joined #openstack-lbaas19:10
*** leitan has quit IRC19:10
*** harlowja has joined #openstack-lbaas19:19
*** slaweq has quit IRC19:19
*** sanfern has quit IRC19:21
*** slaweq has joined #openstack-lbaas19:28
*** longstaff has joined #openstack-lbaas19:36
*** salmankhan has joined #openstack-lbaas19:37
*** longstaff has quit IRC19:41
*** salmankhan has quit IRC19:41
*** jniesz has joined #openstack-lbaas19:45
*** yamamoto has quit IRC19:53
*** longstaff has joined #openstack-lbaas20:02
*** yamamoto has joined #openstack-lbaas20:04
*** yamamoto has quit IRC20:09
*** openstack has quit IRC20:38
*** openstack has joined #openstack-lbaas20:42
*** ChanServ sets mode: +o openstack20:42
*** atoth has quit IRC20:59
*** openstackgerrit has quit IRC21:03
johnsomUgh, so painful the last few weeks21:05
xgerman_+121:05
johnsomTrying to get our openstacksdk gate jobs running again, so these hurt21:05
xgerman_yep…21:06
*** rcernin has joined #openstack-lbaas21:17
*** rcernin has quit IRC21:33
rm_workjohnsom: where was that SP bug? is someone touching that yet?21:39
*** yamamoto has joined #openstack-lbaas21:39
johnsomNo, no one has grabbed it as far as I know.  Just a sec21:39
rm_workk i might look then21:40
johnsomhttps://storyboard.openstack.org/#!/story/200141921:40
rm_workneed that fixed so i can push new code21:40
johnsomrm_work This change broke it: https://review.openstack.org/#/c/458308/66/octavia/controller/worker/tasks/model_tasks.py21:41
rm_worklol so21:42
rm_workthat incomprehensible thing that we all said "hopefully this works"21:42
rm_workdidn't work21:42
johnsomSo, either the SP updates with the old code, or the qos policy update works with the new code21:42
johnsomYeah, pretty much21:42
rm_workdid you figure out WHY the SP doesn't update?21:42
rm_workshouldn't functional tests have caught this actually?21:43
johnsomI think it's a None case that breaks his code21:43
rm_workah :/ k21:43
johnsomNo, it's down in the actual code flow, so a scenario would have had to catch it, which, yeah, is still WIP21:43
johnsomThat is about as far as I got before the break21:44
rm_workerr wait really?21:46
rm_workoh hold on where is this21:46
rm_workthis is below the DB state?21:46
rm_workso the DB will always be correct, just that the generated config is broken?21:46
johnsomThis is not DB, it's updating the data model used in the flow21:47
rm_workerr21:47
*** yamamoto has quit IRC21:47
rm_workright but21:47
rm_workanything between API and DB is testable with functional tests21:47
rm_workactually anything from API to config generation21:47
johnsomRight21:47
rm_workso i can't imagine how a func test couldn't catch this21:47
rm_workshouldn't have to be scenario21:47
johnsomIt's the data model IN the flow21:47
rm_workerr21:48
rm_workoh, so yes, below the DB layer21:48
rm_workmaybe my terminology is bad21:48
rm_workafter the DB21:48
johnsomIt's an update, so this is part of the "handle update rollbacks" so, DB still shows old data, flow data model has new data21:48
rm_workis your previous comment actually right21:50
rm_workabout the issue being the two types passed through?21:50
rm_workor was that just your initial speculation21:50
*** slaweq has quit IRC21:50
johnsominitial speculation.  I think the types end up different, but it's a side effect.21:51
rm_workk21:52
rm_workugh there needs to be a way to build a test for this or fixing it is going to be a nightmare21:52
rm_workwe have a test_model_tasks.py21:53
rm_work...21:53
rm_workalright i'm assigning this to myself and looking into it21:53
johnsomThank you!21:53
*** rcernin has joined #openstack-lbaas22:06
rm_workjohnsom: AFAICT the new code works identically to the old code22:08
rm_workif i pass in an update dict for the pool that includes new SP (either in dict form or object form) the resulting pool object has the new thing in the same way as the old code22:08
johnsomBuild an LB without SP, then update it to have SP or try building with SP and updating.  It bombs22:09
johnsomI reproduced Alex's issue22:09
rm_worki mean22:09
rm_worki'm doing it in unit testing22:09
rm_workaround the UpdateAttributes task22:09
johnsomI understand22:09
rm_workpass in Pool object and update dict22:10
rm_workget out expected results22:10
rm_worki'm not sure how it'd work any differently in practice, if this is the real issue22:10
rm_worki think it must be something else?22:10
johnsomI'm just sharing what I observed.  Like I said, it was crunch time right before the break, so...22:10
johnsomI don't think so, I am pretty sure I flip/flopped this code and resolved the issue but broke  QoS22:11
rm_workhttp://paste.openstack.org/show/641034/22:15
rm_workthis passes22:15
rm_worknot sure how it would if this were the issue22:15
*** longstaff has quit IRC22:22
*** longstaff has joined #openstack-lbaas22:23
*** armax has quit IRC22:40
*** slaweq has joined #openstack-lbaas22:58
*** slaweq has quit IRC23:00
« Sunday, 2018-01-07 Index Tuesday, 2018-01-09 »

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!