Tuesday, 2018-05-29

« Monday, 2018-05-28 Index Wednesday, 2018-05-30 »
*** longkb1 has joined #openstack-lbaas00:36
*** yamamoto has quit IRC02:04
*** yamamoto has joined #openstack-lbaas02:04
*** threestrands has quit IRC02:05
*** threestrands_ has joined #openstack-lbaas02:05
*** raginbajin has quit IRC02:40
*** raginbajin has joined #openstack-lbaas02:40
*** keithmnemonic[m] has quit IRC02:41
*** dayou_ has quit IRC02:41
*** dayou_ has joined #openstack-lbaas02:42
*** oanson has quit IRC02:44
*** PagliaccisCloud has quit IRC02:44
*** lxkong has quit IRC02:44
*** gigo has quit IRC02:44
*** oanson has joined #openstack-lbaas02:45
*** PagliaccisCloud has joined #openstack-lbaas02:45
*** lxkong has joined #openstack-lbaas02:45
*** gigo has joined #openstack-lbaas02:45
*** longkb1 has quit IRC02:46
*** mrodriguez has quit IRC02:46
*** JudeC has quit IRC02:46
*** andreykurilin has quit IRC02:46
*** yamamoto has quit IRC02:46
*** rcernin has quit IRC02:46
*** openstackgerrit has quit IRC02:46
*** thomasem has quit IRC02:46
*** ipsecguy has quit IRC02:46
*** longkb1 has joined #openstack-lbaas02:47
*** mrodriguez has joined #openstack-lbaas02:47
*** JudeC has joined #openstack-lbaas02:47
*** andreykurilin has joined #openstack-lbaas02:47
*** yamamoto has joined #openstack-lbaas02:47
*** rcernin has joined #openstack-lbaas02:47
*** openstackgerrit has joined #openstack-lbaas02:47
*** thomasem has joined #openstack-lbaas02:47
*** ipsecguy has joined #openstack-lbaas02:47
*** imacdonn has quit IRC02:50
*** irenab has quit IRC02:50
*** vegarl has quit IRC02:50
*** isssp has quit IRC02:50
*** phuoc_ has quit IRC02:50
*** pck has quit IRC02:50
*** kbyrne has quit IRC02:50
*** devfaz has quit IRC02:50
*** jmccrory has quit IRC02:50
*** threestrands_ has quit IRC02:50
*** ianychoi_ has quit IRC02:50
*** strigazi has quit IRC02:50
*** Eugene_ has quit IRC02:50
*** ltomasbo has quit IRC02:50
*** imacdonn has joined #openstack-lbaas02:50
*** irenab has joined #openstack-lbaas02:50
*** vegarl has joined #openstack-lbaas02:50
*** isssp has joined #openstack-lbaas02:50
*** threestrands_ has joined #openstack-lbaas02:51
*** ianychoi_ has joined #openstack-lbaas02:51
*** strigazi has joined #openstack-lbaas02:51
*** Eugene_ has joined #openstack-lbaas02:51
*** ltomasbo has joined #openstack-lbaas02:51
*** phuoc_ has joined #openstack-lbaas02:51
*** pck has joined #openstack-lbaas02:51
*** kbyrne has joined #openstack-lbaas02:51
*** devfaz has joined #openstack-lbaas02:51
*** jmccrory has joined #openstack-lbaas02:51
*** annp has joined #openstack-lbaas03:14
*** keithmnemonic[m] has joined #openstack-lbaas03:19
*** hongbin has joined #openstack-lbaas03:28
hongbinjohnsom: ping03:28
johnsomlxkong: yes, we support proxy03:29
johnsomhongbin: hi03:29
hongbinjohnsom: we were talking about the zun octavia integration in the summit, right?03:30
hongbinjohnsom: i might be able to find a volunteer for that03:30
johnsomYes03:30
johnsomOh, that would be great03:30
hongbincool, do you have a BP or storyborad item to track this effort?03:31
johnsomWe have struggled with bugs in nova-lxd, so interested in other options03:31
johnsomWe may have a storyboard for containers, but nothing for Zun specifically.  I am on mobile so can’t look up right now, but feel free to open an RFE for Zun03:32
hongbinsure, i can open one for that and you can review it after03:33
johnsomI am also happy to do a video conference to talk about it and do training on our code03:33
johnsomOk, sounds good!03:34
johnsomI know one of the cores is also interested in this, so you have some good support03:35
hongbinawesome03:44
hongbinhttps://storyboard.openstack.org/#!/story/200211703:44
*** hongbin has quit IRC03:50
*** kiennt26 has joined #openstack-lbaas03:55
*** kiennt26 has quit IRC03:56
*** gans has joined #openstack-lbaas04:10
*** links has joined #openstack-lbaas06:02
*** imacdonn has quit IRC06:16
*** imacdonn has joined #openstack-lbaas06:16
*** pcaruana has joined #openstack-lbaas06:37
*** kobis has joined #openstack-lbaas06:38
*** kobis has quit IRC06:45
*** apple01 has joined #openstack-lbaas07:01
*** dmellado has joined #openstack-lbaas07:12
*** threestrands_ has quit IRC07:14
*** tesseract has joined #openstack-lbaas07:14
*** apple01 has quit IRC07:20
*** apple01 has joined #openstack-lbaas07:23
*** nmanos has joined #openstack-lbaas07:24
*** apple01 has quit IRC07:29
*** kobis has joined #openstack-lbaas07:32
*** kobis has quit IRC07:33
*** kobis has joined #openstack-lbaas07:33
*** apple01 has joined #openstack-lbaas07:39
*** rcernin has quit IRC07:42
*** apple01 has quit IRC07:44
*** apple01 has joined #openstack-lbaas07:44
*** AlexeyAbashkin has joined #openstack-lbaas07:46
*** cgoncalves|pto is now known as cgoncalves07:52
*** numans is now known as numans_afk07:53
*** apple01 has quit IRC08:22
*** apple01 has joined #openstack-lbaas08:29
*** ianychoi_ is now known as ianychoi08:40
*** numans_afk is now known as numans08:53
openstackgerritMichał Dulko proposed openstack/python-octaviaclient master: Add timeout options to listener resource  https://review.openstack.org/57000708:54
*** rcernin has joined #openstack-lbaas08:58
*** salmankhan has joined #openstack-lbaas09:02
*** pcaruana has quit IRC09:23
*** pcaruana has joined #openstack-lbaas09:27
*** kobis has quit IRC09:34
*** pcaruana has quit IRC09:54
*** kobis has joined #openstack-lbaas10:00
*** rcernin has quit IRC10:10
*** yamamoto has quit IRC10:10
*** pcaruana has joined #openstack-lbaas10:10
*** gans has quit IRC10:55
*** yamamoto has joined #openstack-lbaas11:11
*** yamamoto has quit IRC11:16
*** yamamoto has joined #openstack-lbaas11:36
*** pcaruana has quit IRC11:42
*** yamamoto has quit IRC11:47
*** yamamoto has joined #openstack-lbaas11:48
*** leitan has joined #openstack-lbaas11:49
*** atoth has joined #openstack-lbaas11:55
*** pcaruana has joined #openstack-lbaas12:02
*** longkb1 has quit IRC12:16
openstackgerritKobi Samoray proposed openstack/octavia master: Octavia devstack plugin API mode  https://review.openstack.org/57092412:23
*** amuller has joined #openstack-lbaas12:25
*** Alexey_Abashkin has joined #openstack-lbaas12:33
*** AlexeyAbashkin has quit IRC12:36
*** Alexey_Abashkin is now known as AlexeyAbashkin12:36
openstackgerritNoam Manos proposed openstack/octavia-tempest-plugin master: Fixed cleanup long duration  https://review.openstack.org/57001312:37
*** links has quit IRC13:20
*** kobis has quit IRC13:26
*** apple01 has quit IRC13:41
*** apple01 has joined #openstack-lbaas13:42
openstackgerritNoam Manos proposed openstack/octavia-tempest-plugin master: Fixed cleanup long duration  https://review.openstack.org/57001314:17
*** apple01 has quit IRC14:18
*** apple01 has joined #openstack-lbaas14:19
openstackgerritMerged openstack/octavia-dashboard master: Imported Translations from Zanata  https://review.openstack.org/57063314:31
*** apple01 has quit IRC14:32
*** KeithMnemonic has joined #openstack-lbaas14:48
*** nmanos has quit IRC14:55
openstackgerritJan Zerebecki proposed openstack/neutron-lbaas master: Improve speed of loading LBs from DB  https://review.openstack.org/56836115:09
*** apple01 has joined #openstack-lbaas15:09
*** jiteka has joined #openstack-lbaas15:17
*** apple01 has quit IRC15:21
*** apple01 has joined #openstack-lbaas15:21
*** apple01 has quit IRC15:36
*** AlexeyAbashkin has quit IRC15:49
cgoncalvesrm_work, is https://review.openstack.org/#/c/552549 still WIP? what's missing?15:51
*** kobis has joined #openstack-lbaas16:20
openstackgerritMerged openstack/neutron-lbaas master: tempest: healthmonitor is deleted via resource_cleanup  https://review.openstack.org/49212816:20
*** pcaruana has quit IRC16:39
*** JudeC has quit IRC16:53
*** eandersson has quit IRC16:53
*** JudeC has joined #openstack-lbaas17:06
openstackgerritJan Zerebecki proposed openstack/neutron-lbaas master: Improve speed of loading LBs from DB  https://review.openstack.org/56836117:07
*** salmankhan has quit IRC17:12
*** eandersson has joined #openstack-lbaas17:24
*** kobis has quit IRC17:25
*** kobis has joined #openstack-lbaas17:26
*** kobis has quit IRC17:30
rm_workcgoncalves: tests17:39
rm_workcgoncalves: and ... actual testing (I haven't tried it yet)17:39
rm_worki've only just typed up the code, never ran it17:39
rm_workit's just theoretical :P17:39
openstackgerritJan Zerebecki proposed openstack/neutron-lbaas master: Improve speed of loading Pools from DB  https://review.openstack.org/57098617:43
*** pcaruana has joined #openstack-lbaas17:43
*** kobis has joined #openstack-lbaas17:47
rm_workare we *still* trying to fix this?17:50
rm_work^^ DB object loading in n-lbaas17:51
amulleryup, people still use LBaaS v2 with haproxy17:52
amulleron Newton17:52
rm_worki mean how have we not already fixed this17:53
rm_workwe've had like 3-4 patches to address DB loads17:53
amulleryeah17:53
amullerhttps://bugzilla.redhat.com/show_bug.cgi?id=154593917:54
openstackbugzilla.redhat.com bug 1545939 in openstack-neutron-lbaas "neutron lbaas commands take a long time to complete" [Medium,Assigned] - Assigned to cgoncalves17:54
amullerit's a customer bug so almost all of the comments are private17:54
cgoncalvesrm_work, ok. I plan on devstacking it tomorrow and give it a try17:54
amullerbut yeah it seems like the issue was not fixed17:54
rm_work:(17:54
rm_workcgoncalves: ok, let me know -- i mean, i'd like to test it myself, but i really need to get the other tempest testing done first17:55
amullerat least according to this customer on latest RH-OSP 10 (which is Newton + more LBaaS backports after the upstream Newton branched was killed)17:55
cgoncalvesamuller, I've just finished deploying OSP10 (Newton) for that rhbz+patch up for review17:55
cgoncalvesrm_work, the code from Blizzard reduced time in 3-5 seconds, and according to committer upgrading sqlalchemy would drop it down a bit more18:00
rm_workk18:01
openstackgerritCarlos Goncalves proposed openstack/octavia master: Add grenade support  https://review.openstack.org/54965418:14
*** atoth has quit IRC18:20
*** atoth has joined #openstack-lbaas18:21
*** amuller has quit IRC18:25
*** tesseract has quit IRC18:30
rm_workjohnsom: for l7rule testing, do i need to do test_create for every "type"?18:32
rm_work`The L7 rule type. One of COOKIE, FILE_TYPE, HEADER, HOST_NAME, or PATH.`18:32
*** pcaruana has quit IRC18:39
*** phuoc has joined #openstack-lbaas18:44
*** JudeC_ has joined #openstack-lbaas18:47
johnsomHmmm, I would say yes, but that can be an incremental thing18:49
johnsomFor API tests, I haven't done that yet, typically I take one path and make sure I exercise all of the parameters on a call, but I haven't been doing all permutations18:50
*** devfaz_ has joined #openstack-lbaas18:51
*** kbyrne_ has joined #openstack-lbaas18:51
*** jmccrory_ has joined #openstack-lbaas18:51
*** JudeC has quit IRC18:52
*** phuoc_ has quit IRC18:52
*** pck has quit IRC18:52
*** kbyrne has quit IRC18:52
*** devfaz has quit IRC18:53
*** jmccrory has quit IRC18:53
*** kbyrne_ is now known as kbyrne18:53
*** jmccrory_ is now known as jmccrory18:53
*** numans_ has joined #openstack-lbaas18:58
*** numans has quit IRC19:00
rm_workk19:02
*** AlexeyAbashkin has joined #openstack-lbaas19:06
*** numans_ has quit IRC19:10
*** AlexeyAbashkin has quit IRC19:10
*** AlexeyAbashkin has joined #openstack-lbaas19:11
*** samccann has joined #openstack-lbaas19:12
*** salmankhan has joined #openstack-lbaas19:16
*** numans has joined #openstack-lbaas19:16
*** salmankhan has quit IRC19:20
*** AlexeyAbashkin has quit IRC19:34
*** kobis has quit IRC19:37
rm_workjohnsom: umm... if i have one l7policy on a listener, created in position 1 (which makes sense)19:56
rm_workand then I update that policy to be "position: 2"19:56
rm_workdo we not re-calculate all of the positions and renumber starting from 1?19:57
rm_workthus it ends up as position = 1 anyway?19:57
rm_work(what I'm seeing is that I do a GET and it still shows position 2, which seems awkward)19:57
johnsomUmm, I think we do. This whole ordering thing is a bit "odd" to me.  Stephen listed it out though here: https://docs.openstack.org/octavia/latest/user/guides/l7.html#policy-position19:57
rm_worki mean I THOUGHT it made sense to me19:58
rm_workbut either my understanding was wrong, or we're doing it wrong in octavia19:58
rm_worksbalukoff: advise plz :P19:58
rm_workhmm the haproxy ordering thing is kinda lame19:59
rm_workbasically position hardly matters if you have a mix of stuff19:59
*** harlowja has joined #openstack-lbaas20:00
rm_workso it's impossible to be like ... redirect to url if X, to pool if Y, otherwise reject everything else20:00
johnsomI don't follow, why can't you do that?20:01
rm_worksee the link you just posted20:01
rm_workREJECT policies take precedence over all other policies.20:02
rm_workREDIRECT_TO_URL policies take precedence over REDIRECT_TO_POOL policies.20:02
rm_workREDIRECT_TO_POOL policies are only evaluated after all of the above, and in the order specified by the position of the policy.20:02
rm_workso you can't "REJECT *" in last position after your pool/url redirects20:02
johnsomHmm, wonder if that is still the case20:03
johnsomAh, it's just how the jinja is laid out20:09
rm_workorly20:11
rm_workwell that's ... in need of a rewrok20:11
rm_work*rework20:11
rm_workneed a story for that IMO20:11
rm_workI'll go make one20:11
rm_workjohnsom: but umm... if you have a sec, see if you can find any code in octavia that would do policy ordering20:12
rm_workbecause i can't find ANYTHING20:12
rm_worki'm not sure we even enforce simple things like "two can't have the same order"20:12
rm_workmaybe we only did the fancy reordering code in n-lbaas?20:12
rm_workhmm, seems that creating it in position 2, it does go to position 120:17
*** leitan has quit IRC20:17
*** leitan has joined #openstack-lbaas20:17
rm_workand inserting works too20:18
rm_workok20:18
rm_workso i must just not be seeing it...20:18
johnsomSorry, switched windows and didn't hear the bong20:18
johnsomhttps://github.com/openstack/octavia/blob/master/octavia/common/jinja/haproxy/templates/macros.j2#L10120:18
rm_worknp i got more info20:18
johnsomSo, that if conditional would enforce the order listed in the doc20:19
johnsomIt does seem like this could be done better, just some effort20:20
*** AlexeyAbashkin has joined #openstack-lbaas20:20
rm_workhmm tested manually and the update seems to be calculating properly20:20
rm_worknot sure wtf20:20
*** leitan has quit IRC20:22
*** AlexeyAbashkin has quit IRC20:25
rm_workjohnsom: err looking at that macro, i don't see why it'd enforce any order besides position20:26
rm_workit's just going through the list in-order, so long as the list is sorted by position...20:26
rm_workno?20:27
rm_workerr wait, would have to look *above* this20:27
rm_workah yeah, 132-13520:27
rm_workjust looping over the policies and adding them in order20:27
johnsomSo, yeah, ok, maybe the doc is wrong?20:28
rm_workmaybe20:28
rm_workwhat I assumed was that it had to do with haproxy itself20:28
rm_workOH the update issue I had was because of the cleanup order stuff20:31
*** samccann has quit IRC20:53
openstackgerritAdam Harwell proposed openstack/octavia-tempest-plugin master: Create api+scenario tests for l7policies  https://review.openstack.org/57048220:57
openstackgerritAdam Harwell proposed openstack/octavia-tempest-plugin master: Create api+scenario tests for l7rules  https://review.openstack.org/57048520:57
rm_workjohnsom: ^^ ok that should be looking better, at least for API tests20:58
rm_workoh whoops pep820:58
johnsomNice, I will add the depends-on20:58
rm_workfixing <_<20:59
johnsomKobi also raised an issue this morning, so I hope to do a cleanup patch spin soon. Still working on summit wrapup stuff20:59
rm_workkk20:59
openstackgerritAdam Harwell proposed openstack/octavia-tempest-plugin master: Create api+scenario tests for l7policies  https://review.openstack.org/57048221:08
openstackgerritAdam Harwell proposed openstack/octavia-tempest-plugin master: Create api+scenario tests for l7rules  https://review.openstack.org/57048521:08
rm_workok there we go21:08
openstackgerritMichael Johnson proposed openstack/octavia master: Implement provider drivers - L7 Policy  https://review.openstack.org/56705921:08
openstackgerritMichael Johnson proposed openstack/octavia master: Implement provider drivers - L7 Rules  https://review.openstack.org/56707321:09
openstackgerritJan Zerebecki proposed openstack/neutron-lbaas master: Remove eager loading of Listener relations  https://review.openstack.org/57059621:18
openstackgerritJan Zerebecki proposed openstack/neutron-lbaas master: Remove eager loading of L7Policy relations  https://review.openstack.org/57102621:22
*** sshank has joined #openstack-lbaas21:44
*** rcernin has joined #openstack-lbaas22:18
johnsomrm_work Did we talk about this during the summit? https://storyboard.openstack.org/#!/story/200211422:28
johnsomListener not going ONLINE?22:28
rm_workumm22:28
rm_workthat usually means the HMs aren't *actually* working22:29
rm_workis that in his cloud?22:29
johnsomRight22:29
rm_workI've run all gates (including API tests) as live tests in my devstack and they pass22:29
johnsomI think that is the case here, that the HM isn't getting it's heartbeat.  I just vaguely remember something like this (probably vexxhost debug before the summit)22:29
rm_workyeah22:29
johnsomYeah, me too, I always run live22:29
rm_worki'll ask for some details22:30
johnsomYeah, I will comment22:30
rm_workwait for mine22:31
rm_worki don't want to tip off something22:31
johnsomHa, ok22:31
rm_workoops, i didn't test in *noop* for the policy ones22:56
rm_worklol22:56
*** leitan has joined #openstack-lbaas22:59
johnsomThis is an interesting one: https://storyboard.openstack.org/#!/story/200211823:07
johnsomI need to see the haproxy config to see where that went off the rails23:07
johnsomOh, nevermind...23:10
johnsomI think that is an haproxy 1.5 on centos issue23:10
KeithMnemonicjohnson: is there a getting started guide for octavia on AIO or was that what I was on the hook for?23:10
KeithMnemonicjohnsonm: is there a getting started guide for octavia on AIO or was that what I was on the hook for?23:10
KeithMnemonicArgh hate when I type too fast. johnsom: is there a getting started guide for octavia on AIO or was that what I was on the hook for?23:11
johnsomKeithMnemonic Hey Keith, we have this: https://docs.openstack.org/octavia/latest/install/index.html23:11
johnsomBut it is very high level. The detailed guide was what we were going to work together one23:11
johnsomon23:11
KeithMnemonicok and nothing for AIO? so if i hack on it i should take notes ;-)23:12
johnsomAIO? Like Ansible?23:12
johnsomOr devstack?23:12
KeithMnemonicyeah ansible all in one. I thought xgerman_ once mentioned using AIO to play with it23:13
johnsomYeah, hang on, I think there is a guide over in the Ansible repo23:15
johnsomKeithMnemonic https://docs.openstack.org/openstack-ansible-os_octavia/latest/23:15
KeithMnemonicthat is it !! but you want to still work on a getting started guide that will be applicable to devstack and/or AIO?23:17
johnsomI have started, but not yet posted a detailed install guide for doing it by hand.23:18
johnsomThat is what we need, but is a bunch of work that keeps getting bumped down the priority stack.23:19
*** leitan has quit IRC23:21
rm_workjohnsom: trying to figure out how the heck the policies went online in this noop: http://logs.openstack.org/82/570482/3/check/octavia-v2-dsvm-noop-api/b62f93c/testr_results.html.gz23:21
rm_workwe do seem to create them as OFFLINE in the API...23:22
*** leitan has joined #openstack-lbaas23:22
johnsomHmm, yeah, I think those are faked up in the amp agent23:22
rm_workwut23:23
johnsomPolicies and rules don't really have a status in haproxy23:24
rm_workhmmm23:26
rm_workyeah i suppose not23:26
rm_workso need to find where we set them to ONLINE23:26
johnsomhttps://github.com/openstack/octavia/blob/master/octavia/controller/worker/tasks/database_tasks.py#L179423:26
*** leitan has quit IRC23:26
rm_workahhh yeah i just found it too23:27
rm_worklol k23:27
rm_workfixed23:31
openstackgerritAdam Harwell proposed openstack/octavia-tempest-plugin master: Create api+scenario tests for l7policies  https://review.openstack.org/57048223:31
openstackgerritAdam Harwell proposed openstack/octavia-tempest-plugin master: Create api+scenario tests for l7rules  https://review.openstack.org/57048523:31
*** leitan has joined #openstack-lbaas23:32
*** eandersson has quit IRC23:32
*** openstackstatus has joined #openstack-lbaas23:43
*** ChanServ sets mode: +v openstackstatus23:43
« Monday, 2018-05-28 Index Wednesday, 2018-05-30 »

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