| *** mlavalle has quit IRC | 00:02 | |
| *** sbfox has quit IRC | 00:04 | |
| *** sbfox has joined #openstack-neutron | 00:05 | |
| *** digambar has quit IRC | 00:06 | |
| *** banix has joined #openstack-neutron | 00:09 | |
| *** alagalah has quit IRC | 00:17 | |
| *** armax has quit IRC | 00:17 | |
| *** Jabadia_ has joined #openstack-neutron | 00:20 | |
| *** Jabadia_ has quit IRC | 00:24 | |
| *** thesanchez has joined #openstack-neutron | 00:26 | |
| *** matsuhashi has joined #openstack-neutron | 00:27 | |
| *** sbfox has quit IRC | 00:29 | |
| *** qs201 has joined #openstack-neutron | 00:31 | |
| *** sandr8 has quit IRC | 00:34 | |
| *** BuSerD has quit IRC | 00:34 | |
| *** banix has quit IRC | 00:37 | |
| *** banix has joined #openstack-neutron | 00:38 | |
| *** manishg has quit IRC | 00:39 | |
| *** xuhanp has joined #openstack-neutron | 00:39 | |
| *** WackoRobie has joined #openstack-neutron | 00:40 | |
| *** markmcclain has quit IRC | 00:45 | |
| *** tomoe_ has joined #openstack-neutron | 00:45 | |
| *** BuSerD has joined #openstack-neutron | 00:47 | |
| *** changbl has joined #openstack-neutron | 00:48 | |
| *** manishg has joined #openstack-neutron | 00:49 | |
| *** markmcclain has joined #openstack-neutron | 00:53 | |
| *** _cjones_ has quit IRC | 00:56 | |
| *** _cjones_ has joined #openstack-neutron | 00:57 | |
| *** Sukhdev has quit IRC | 00:58 | |
| *** zigo has quit IRC | 00:59 | |
| *** crc32 has quit IRC | 00:59 | |
| *** zigo has joined #openstack-neutron | 01:01 | |
| *** _cjones_ has quit IRC | 01:01 | |
| openstackgerrit | Ronak Shah proposed a change to openstack/neutron: Nuage Plugin: Delete router requires precommit checks https://review.openstack.org/84304 | 01:03 |
|---|---|---|
| *** oda-g has quit IRC | 01:03 | |
| *** jlibosva has joined #openstack-neutron | 01:06 | |
| *** xuhanp has quit IRC | 01:08 | |
| *** banix has quit IRC | 01:10 | |
| *** tomoe_ has quit IRC | 01:10 | |
| *** networkstatic is now known as networkstatic_zZ | 01:10 | |
| *** tomoe_ has joined #openstack-neutron | 01:10 | |
| *** jlibosva has quit IRC | 01:10 | |
| *** tomoe_ has quit IRC | 01:15 | |
| *** Jabadia has joined #openstack-neutron | 01:21 | |
| *** networkstatic_zZ is now known as networkstatic | 01:22 | |
| *** SumitNaiksatam has joined #openstack-neutron | 01:23 | |
| *** manishg has quit IRC | 01:23 | |
| *** Jabadia has quit IRC | 01:25 | |
| *** alagalah has joined #openstack-neutron | 01:28 | |
| *** beagles has quit IRC | 01:30 | |
| *** jecarey has joined #openstack-neutron | 01:30 | |
| *** alagalah has quit IRC | 01:32 | |
| *** manishg has joined #openstack-neutron | 01:33 | |
| *** manishg_ has joined #openstack-neutron | 01:34 | |
| *** manishg has quit IRC | 01:38 | |
| *** manishg_ is now known as manishg | 01:38 | |
| *** hemanthravi has joined #openstack-neutron | 01:40 | |
| *** xuhanp has joined #openstack-neutron | 01:47 | |
| *** sbfox has joined #openstack-neutron | 01:49 | |
| *** arosen1 has quit IRC | 01:49 | |
| openstackgerrit | berlin proposed a change to openstack/neutron: Check NVP router's status before deploying a service https://review.openstack.org/84072 | 02:01 |
| HenryG | kevinbenton: ping | 02:06 |
| kevinbenton | henryg: pong | 02:06 |
| HenryG | kevinbenton: you are with big switch right? | 02:07 |
| kevinbenton | henryg: yep | 02:07 |
| HenryG | kevinbenton: can you look at https://bugs.launchpad.net/neutron/+bug/1298459 | 02:07 |
| *** jlibosva has joined #openstack-neutron | 02:07 | |
| HenryG | kevinbenton: I can help you figure it out (and why it is showing up now) | 02:07 |
| HenryG | kevinbenton: it is related to https://review.openstack.org/40296 | 02:08 |
| kevinbenton | henryg: i see. i think i got it. will you be online for a little while? | 02:09 |
| HenryG | kevinbenton: another 2 hours or so | 02:09 |
| *** xianghui has joined #openstack-neutron | 02:09 | |
| *** flwang has quit IRC | 02:12 | |
| HenryG | arosen: ping | 02:12 |
| *** jlibosva has quit IRC | 02:13 | |
| mestery | HenryG: Making the rounds tonight it appears. :) | 02:13 |
| HenryG | mestery: yup, hunting those affected by https://review.openstack.org/40296 :) | 02:14 |
| mestery | HenryG: Nice work, and thanks for covering this! | 02:14 |
| HenryG | mestery: do you know any embrane nicks? | 02:15 |
| *** sbfox has quit IRC | 02:15 | |
| mestery | Ivar is the Embrane dude, but he doesn't appear to be online at the moment. | 02:16 |
| HenryG | mestery: thanks! | 02:17 |
| mestery | HenryG: Anytime. ;) | 02:18 |
| *** banix has joined #openstack-neutron | 02:19 | |
| *** jrbudnack has quit IRC | 02:20 | |
| *** Jabadia has joined #openstack-neutron | 02:22 | |
| *** dims has quit IRC | 02:23 | |
| *** dims has joined #openstack-neutron | 02:24 | |
| *** banix has quit IRC | 02:24 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Add missing DB migrations for BSN ML2 and plugin https://review.openstack.org/84313 | 02:25 |
| kevinbenton | HenryG: ^^ | 02:25 |
| kevinbenton | i think that should cover it | 02:25 |
| HenryG | kevinbenton: hmmm | 02:27 |
| *** Jabadia has quit IRC | 02:27 | |
| kevinbenton | HenryG: I’m slightly concerned about the fact that I’m revising the ‘icehouse’ revision | 02:29 |
| kevinbenton | HenryG: what is the signifigance of that revision? our plugin and driver won’t be broken in icehouse, will it? | 02:29 |
| HenryG | kevinbenton: problem is that 40296 breaks migrations earlier than icehouse | 02:30 |
| *** manishg has quit IRC | 02:30 | |
| *** amotoki_ has quit IRC | 02:30 | |
| HenryG | kevinbenton: so I think we'll need to go and do some fiddling with existing migration scripts | 02:30 |
| kevinbenton | HenryG: is that the preferred approach? | 02:31 |
| HenryG | kevinbenton: let's see if amotoki is online, he can offer some advice | 02:31 |
| HenryG | kevinbenton: let's start with the consistencyhashes table | 02:32 |
| HenryG | kevinbenton: that one is OK for the bsn plugin | 02:32 |
| HenryG | kevinbenton: for ml2, I think it simply needs to be added to ml2_initial | 02:33 |
| kevinbenton | HenryG: i think it would just make more sense to add it to the same consistency hash migration script for the plugin if we are editing previous scripts | 02:33 |
| kevinbenton | HenryG: but I was under the impression that editing existing scripts is blasphemy :-) | 02:34 |
| HenryG | kevinbenton: when did your mech driver get added to ml2? (what migration?) | 02:34 |
| kevinbenton | befor the consistency hash functionality was added | 02:35 |
| kevinbenton | before* | 02:35 |
| HenryG | kevinbenton: 40296 requires some editing of some past transgressions (existing migration scripts) | 02:36 |
| HenryG | kevinbenton: ok, then I think we can just add ML2 to the list of plugins in 81c553f3776c_bsn_consistencyhashes.py | 02:38 |
| HenryG | kevinbenton: so here is how you test this: | 02:38 |
| HenryG | Clone neutron master and apply 40296 | 02:39 |
| HenryG | Run devstack with ML2 + your mech driver | 02:41 |
| HenryG | You should see missing table error | 02:41 |
| HenryG | Unstack and add ML2 to 81c553f3776c_bsn_consistencyhashes.py | 02:42 |
| HenryG | Run devstack again and table should no longer be missing | 02:43 |
| HenryG | OK? | 02:44 |
| HenryG | Now for the networkdhcpagentbindings table in the legacy plugin | 02:44 |
| HenryG | For that one I think you just need to add your plugin to 4692d074d587_agent_scheduler.py | 02:45 |
| HenryG | Test in the same way | 02:45 |
| HenryG | kevinbenton: ^^^ make sense? | 02:46 |
| kevinbenton | HenryG: can we get confirmation from another core that editing a previous test is ok? | 02:51 |
| *** hemanthravi has quit IRC | 02:51 | |
| kevinbenton | HenryG: I’m just going to cheat and let the BSN automated testing run this stuff for me by setting a dependency in my patch on 40296 | 02:52 |
| HenryG | kevinbenton: absolutely! amotoki, markmcclain and salv-orlando are good candidates | 02:52 |
| HenryG | kevinbenton: that's not cheating, that's being smart | 02:52 |
| kevinbenton | HenryG: wow, it just occurred to me how old this patch was. why wait until now? | 02:53 |
| HenryG | salv-orlando started it and got bored ;) then gave it to me when I opened my big mouth :O | 02:54 |
| kevinbenton | ah, makes sense :-) | 02:56 |
| *** manishg has joined #openstack-neutron | 02:56 | |
| HenryG | Many many plugins break from this patch, not just big switch. Also cisco, nsx, embrane in addition to ml2, ovs, lb, hyperv ... | 02:57 |
| *** markmcclain has quit IRC | 02:59 | |
| *** harlowja is now known as harlowja_away | 02:59 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Add missing DB migrations for BSN ML2 plugin https://review.openstack.org/84313 | 03:00 |
| kevinbenton | HenryG: ^^ that one is definitely easier on the eyes :-) | 03:01 |
| HenryG | kevinbenton: cool! When your CI reports, look for "ProgrammingError: (ProgrammingError) (1146," in the logs. | 03:02 |
| kevinbenton | well if it passes, we know it worked, right? | 03:03 |
| kevinbenton | these missing tables should break things | 03:03 |
| HenryG | true | 03:03 |
| *** jecarey_ has joined #openstack-neutron | 03:07 | |
| HenryG | kevinbenton: looking at the last bsn CI log from 40296 ... it seems the subnetroutes table is also missing :( | 03:07 |
| HenryG | I am updating the bug | 03:08 |
| *** jlibosva has joined #openstack-neutron | 03:09 | |
| *** harlowja_away is now known as harlowja | 03:09 | |
| *** jecarey has quit IRC | 03:10 | |
| kevinbenton | link, i didn’t think we used that table | 03:10 |
| kevinbenton | ? | 03:10 |
| *** chandan_kumar has joined #openstack-neutron | 03:11 | |
| kevinbenton | is it the ML2 plugin or the Big Switch plugin that complained? | 03:11 |
| *** ramishra has joined #openstack-neutron | 03:12 | |
| HenryG | http://openstack-ci-gw.bigswitch.com/logs/refs-changes-96-40296-13/BSN_PLUGIN/logs/screen/screen-q-svc.log.gz | 03:12 |
| HenryG | not ml2 | 03:12 |
| *** yamahata has joined #openstack-neutron | 03:13 | |
| *** jlibosva has quit IRC | 03:13 | |
| *** ramishra_ has joined #openstack-neutron | 03:18 | |
| *** tomoe_ has joined #openstack-neutron | 03:18 | |
| *** ramishra has quit IRC | 03:21 | |
| *** matsuhashi has quit IRC | 03:22 | |
| *** Jabadia has joined #openstack-neutron | 03:22 | |
| *** jecarey__ has joined #openstack-neutron | 03:25 | |
| *** Jabadia has quit IRC | 03:27 | |
| *** tomoe_ has quit IRC | 03:29 | |
| *** jecarey_ has quit IRC | 03:29 | |
| *** tomoe_ has joined #openstack-neutron | 03:30 | |
| *** flwang has joined #openstack-neutron | 03:30 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Add missing DB migrations for BSN ML2 plugin https://review.openstack.org/84313 | 03:30 |
| kevinbenton | HenryG: interesting. i learned something new today. :-) subnet routes aren’t part of a normal subnet | 03:31 |
| *** manishg_ has joined #openstack-neutron | 03:32 | |
| *** tomoe_ has quit IRC | 03:34 | |
| *** chandan_kumar has quit IRC | 03:35 | |
| *** manishg has quit IRC | 03:35 | |
| *** manishg_ is now known as manishg | 03:35 | |
| HenryG | kevinbenton: I did not know that either :-) | 03:38 |
| HenryG | kevinbenton: you find the right migration to fix too, good job | 03:39 |
| *** alexpilotti has joined #openstack-neutron | 03:40 | |
| *** lollipop has joined #openstack-neutron | 03:40 | |
| *** banix has joined #openstack-neutron | 03:42 | |
| *** zhipeng has joined #openstack-neutron | 03:43 | |
| *** tomoe_ has joined #openstack-neutron | 03:43 | |
| *** ramishra_ has quit IRC | 03:44 | |
| *** ramishra has joined #openstack-neutron | 03:45 | |
| *** markwash has quit IRC | 03:47 | |
| *** zhipeng has quit IRC | 03:47 | |
| *** BuSerD has quit IRC | 03:48 | |
| *** markwash has joined #openstack-neutron | 03:50 | |
| *** alexpilotti has quit IRC | 03:52 | |
| *** tomoe_ has quit IRC | 03:57 | |
| *** tomoe_ has joined #openstack-neutron | 03:58 | |
| *** chandan_kumar has joined #openstack-neutron | 03:59 | |
| *** chandan_kumar has quit IRC | 04:01 | |
| kevinbenton | HenryG: had to adjust our CI script. by default it ignores patches with migrations that don’t change the core neutron behavior or the plugin | 04:02 |
| *** tomoe_ has quit IRC | 04:02 | |
| HenryG | kevinbenton: ok. Hope it works out. I am off to bed. I won't be around much tomorrow but will try to check in. Back on Wednesday. | 04:05 |
| *** chandan_kumar has joined #openstack-neutron | 04:06 | |
| *** WackoRobie has quit IRC | 04:06 | |
| *** WackoRobie has joined #openstack-neutron | 04:07 | |
| *** tomoe_ has joined #openstack-neutron | 04:09 | |
| *** jlibosva has joined #openstack-neutron | 04:10 | |
| *** WackoRobie has quit IRC | 04:11 | |
| *** thuc has joined #openstack-neutron | 04:13 | |
| *** thuc_ has joined #openstack-neutron | 04:13 | |
| *** matsuhashi has joined #openstack-neutron | 04:14 | |
| *** jlibosva has quit IRC | 04:15 | |
| *** thuc has quit IRC | 04:17 | |
| *** tomoe_ has quit IRC | 04:20 | |
| *** thuc has joined #openstack-neutron | 04:21 | |
| *** Jabadia has joined #openstack-neutron | 04:23 | |
| *** thuc_ has quit IRC | 04:24 | |
| *** tomoe_ has joined #openstack-neutron | 04:27 | |
| *** Jabadia has quit IRC | 04:28 | |
| *** yfried__ has joined #openstack-neutron | 04:36 | |
| *** networkstatic has quit IRC | 04:38 | |
| openstackgerrit | Ian Wienand proposed a change to openstack/neutron: Record and log reason for dhcp agent resync https://review.openstack.org/81173 | 04:47 |
| *** manishg has quit IRC | 04:48 | |
| *** sandr8 has joined #openstack-neutron | 04:50 | |
| *** banix has quit IRC | 04:50 | |
| *** rand738 has quit IRC | 04:52 | |
| kevinbenton | akihiro: ping | 04:52 |
| amotoki | kevinbenton: pong | 04:52 |
| kevinbenton | amotoki: whoops, wrong username. :-) I’m working on the migration stuff and it looks like a portbindings one might be missing for ML2 | 04:53 |
| *** rand738 has joined #openstack-neutron | 04:53 | |
| kevinbenton | amotoki: is there already a bug for that one? | 04:54 |
| amotoki | kevinbenton: yeah, disabling auto-generation reveals many issues :-) AFAIK, we have no bug for it. | 04:54 |
| *** WackoRobie has joined #openstack-neutron | 04:55 | |
| amotoki | kevinbenton: i wonder why jenkins with ML2 works... | 04:55 |
| kevinbenton | amotoki: i’m bringing up a testbed instance to see why | 04:56 |
| *** sbfox has joined #openstack-neutron | 05:01 | |
| *** matsuhas_ has joined #openstack-neutron | 05:02 | |
| *** matsuhas_ has quit IRC | 05:02 | |
| *** matsuhashi has quit IRC | 05:05 | |
| *** ramishra has quit IRC | 05:05 | |
| *** matsuhashi has joined #openstack-neutron | 05:05 | |
| *** WackoRobie has quit IRC | 05:06 | |
| *** ramishra has joined #openstack-neutron | 05:06 | |
| *** harlowja is now known as harlowja_away | 05:11 | |
| *** jlibosva has joined #openstack-neutron | 05:11 | |
| *** thuc has quit IRC | 05:13 | |
| *** thuc has joined #openstack-neutron | 05:14 | |
| *** jlibosva has quit IRC | 05:16 | |
| *** Jabadia has joined #openstack-neutron | 05:18 | |
| *** thuc has quit IRC | 05:18 | |
| *** qs201 has quit IRC | 05:21 | |
| *** irenab has joined #openstack-neutron | 05:23 | |
| openstackgerrit | Yaguang Tang proposed a change to openstack/python-neutronclient: Make neutronclient parse keystone v3 endpoints correctly https://review.openstack.org/81767 | 05:25 |
| *** sbfox has quit IRC | 05:27 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Add missing DB migrations for BSN ML2 plugin https://review.openstack.org/84313 | 05:29 |
| kevinbenton | amotoki: ^^ | 05:29 |
| kevinbenton | amotoki: that clears up the plugin side for bigswitch | 05:30 |
| amotoki | kevinbenton: nice! | 05:30 |
| kevinbenton | amotoki: i will submit another for that ml2 portbindings table | 05:30 |
| kevinbenton | amotoki: then the bigswitch ml2 driver will work too | 05:30 |
| *** otherwiseguy has quit IRC | 05:31 | |
| kevinbenton | amotoki: i’m not sure why default jenkins passes. the table definitely isn’t added | 05:31 |
| *** sbfox has joined #openstack-neutron | 05:31 | |
| kevinbenton | amotoki: maybe the openvswitch driver never references the portbindings stuff | 05:31 |
| amotoki | kevinbenton: ovs agent MD might not use portbinding table... we need to check it. | 05:32 |
| amotoki | kevinbenton: it looks strange, but ... | 05:32 |
| kevinbenton | amotoki: yeah, driver doesn’t reference anything | 05:33 |
| kevinbenton | https://github.com/bigswitch/neutron/blob/master/neutron/plugins/ml2/drivers/mech_openvswitch.py | 05:33 |
| *** Longgeek has joined #openstack-neutron | 05:35 | |
| *** rotbeard has joined #openstack-neutron | 05:37 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Add ML2 plugin to portbinding migration script https://review.openstack.org/84354 | 05:42 |
| kevinbenton | amotoki: is it possible to merge this, or is this something that will have to wait until after icehouse? https://review.openstack.org/#/c/83855/ | 05:43 |
| kevinbenton | amotoki: just some company name references. no functional changes | 05:43 |
| amotoki | kevinbenton: IMO it looks okay. it is related to branding topic and it is better to be in icehouse. | 05:47 |
| kevinbenton | amotoki: thanks | 05:48 |
| *** pradipta_away is now known as pradipta | 05:54 | |
| openstackgerrit | Liping Mao proposed a change to openstack/neutron: lbaas on a network without gateway https://review.openstack.org/84058 | 05:56 |
| *** sandr8 has quit IRC | 05:56 | |
| *** WackoRobie has joined #openstack-neutron | 06:01 | |
| *** sbfox has quit IRC | 06:02 | |
| *** sbfox has joined #openstack-neutron | 06:03 | |
| *** WackoRobie has quit IRC | 06:05 | |
| *** sandr8 has joined #openstack-neutron | 06:06 | |
| *** jlibosva has joined #openstack-neutron | 06:12 | |
| *** djoreilly has joined #openstack-neutron | 06:13 | |
| openstackgerrit | Liping Mao proposed a change to openstack/neutron: lbaas on a network without gateway https://review.openstack.org/84058 | 06:13 |
| *** jlibosva has quit IRC | 06:17 | |
| *** WackoRobie has joined #openstack-neutron | 06:18 | |
| *** saju_m has joined #openstack-neutron | 06:21 | |
| *** WackoRobie has quit IRC | 06:23 | |
| *** thuc has joined #openstack-neutron | 06:24 | |
| *** sridhar has joined #openstack-neutron | 06:25 | |
| openstackgerrit | Pierre Rambaud proposed a change to openstack/neutron: Fix quota_health_monitor opt name in neutron.conf https://review.openstack.org/84003 | 06:26 |
| *** thuc has quit IRC | 06:29 | |
| openstackgerrit | Zang MingJie proposed a change to openstack/neutron: db migration failed when plugin name alias used in config file https://review.openstack.org/81221 | 06:30 |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/84366 | 06:31 |
| openstackgerrit | dan wendlandt proposed a change to openstack/neutron: Remove last parts of Quantum compatibility shim https://review.openstack.org/84367 | 06:31 |
| *** sungju_ has quit IRC | 06:31 | |
| *** Jabadia_ has joined #openstack-neutron | 06:35 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: BigSwitch: Semaphore on port status update https://review.openstack.org/84368 | 06:35 |
| *** Jabadia has quit IRC | 06:35 | |
| *** Jabadia_ has quit IRC | 06:35 | |
| *** Jabadia has joined #openstack-neutron | 06:36 | |
| *** gdubreui has quit IRC | 06:36 | |
| *** tomoe_ has quit IRC | 06:43 | |
| *** tomoe_ has joined #openstack-neutron | 06:43 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Restore NOT NULL constraint lost by earlier migrations https://review.openstack.org/78591 | 06:44 |
| openstackgerrit | Liping Mao proposed a change to openstack/neutron: lbaas on a network without gateway https://review.openstack.org/84058 | 06:45 |
| *** tomoe_ has quit IRC | 06:47 | |
| openstackgerrit | Liping Mao proposed a change to openstack/neutron: lbaas on a network without gateway https://review.openstack.org/84058 | 06:55 |
| *** matsuhashi has quit IRC | 06:58 | |
| kevinbenton | amotoki: ping | 06:58 |
| *** matsuhashi has joined #openstack-neutron | 06:58 | |
| openstackgerrit | Yaguang Tang proposed a change to openstack/python-neutronclient: Make neutronclient parse keystone v3 endpoints correctly https://review.openstack.org/81767 | 06:59 |
| amotoki | kevinbenton: pong | 06:59 |
| kevinbenton | amotoki: i think the error must have been related to this https://github.com/bigswitch/neutron/blob/master/neutron/db/migration/migrate_to_ml2.py#L360 | 07:00 |
| *** jistr has joined #openstack-neutron | 07:02 | |
| kevinbenton | amotoki: might not be an actual migration problem for ML2. instead it’s just related to where my patch was based | 07:02 |
| *** matsuhashi has quit IRC | 07:03 | |
| amotoki | Was portbinding table changed after the script was merge? IIRC bob refactors portbindings code after BSN ML2 merge. | 07:03 |
| *** evgenyf has joined #openstack-neutron | 07:03 | |
| amotoki | kevinbenton: what test hits this error? grenade? | 07:04 |
| kevinbenton | amotoki: it was happening when just running devstack with the BSN ML2 plugin | 07:05 |
| *** matsuhashi has joined #openstack-neutron | 07:06 | |
| kevinbenton | amotoki: s/plugin/driver. | 07:06 |
| kevinbenton | amotoki: and I don’t think our driver references the portbindings table | 07:06 |
| amotoki | kevinbenton: but this is the script to migrate to OVS plugin to ML2 ovs MD. I wonder why this hits BSN ML2 devstack. | 07:07 |
| kevinbenton | amotoki: well something is still referencing the portbindingports table, and it’s either not being initially created or it’s being deleted. i’m not sure which | 07:09 |
| *** sbfox has quit IRC | 07:10 | |
| kevinbenton | amotoki: because it results in this error http://openstack-ci-gw.bigswitch.com/logs/refs-changes-13-84313-3/BSN_ML2/logs/screen/screen-q-svc.log.gz | 07:10 |
| *** jlibosva has joined #openstack-neutron | 07:12 | |
| amotoki | kevinbenton: ML2 plugin uses different tables for portbindings: ml2_port_bindings | 07:14 |
| kevinbenton | amotoki: yeah, that’s why i’m surprised by the reference to the portbindings table | 07:16 |
| *** luqas has joined #openstack-neutron | 07:16 | |
| *** jlibosva has quit IRC | 07:17 | |
| amotoki | kevinbenton: yeah.. that's suprising indeed. | 07:19 |
| *** WackoRobie has joined #openstack-neutron | 07:19 | |
| kevinbenton | amotoki: i might just wait for the other patch fixing the missing BSN tables to merge and for 40296 to be rebased before worrying about it | 07:22 |
| amotoki | kevinbenton: does bsn ml2 MD load bsn plugin? | 07:22 |
| amotoki | kevinbenton: bsn plugin loads db.portbindings via db.porttracker as far as i understand. | 07:22 |
| kevinbenton | amotoki: oh, that’s precisely the cause then | 07:23 |
| amotoki | kevinbenton: if db.portbindings is loaded, portbindingports registers dict_extend_funcs automatically. | 07:23 |
| *** WackoRobie has quit IRC | 07:24 | |
| *** devvesa has joined #openstack-neutron | 07:24 | |
| kevinbenton | amotoki: i forgot the port tracking was leveraging the port bindings table | 07:25 |
| openstackgerrit | Zang MingJie proposed a change to openstack/neutron: db migration failed when plugin name alias used in config file https://review.openstack.org/81221 | 07:26 |
| amotoki | kevinbenton: it seems you find the cause. db migration reveals much :-) & :-( | 07:26 |
| *** leseb has joined #openstack-neutron | 07:28 | |
| amotoki | kevinbenton: i usually check the difference between after db migration and after neutron-server with schema generation by dumping schema. | 07:30 |
| *** salv-orlando has quit IRC | 07:30 | |
| *** salv-orlando has joined #openstack-neutron | 07:30 | |
| kevinbenton | amotoki: makes sense. i will need to fix this before that patch can go in | 07:32 |
| kevinbenton | amotoki: in the mean time, can you take a look at this? https://review.openstack.org/#/c/84368/ | 07:33 |
| kevinbenton | amotoki: i found another deadlock hit when i was running tests | 07:33 |
| kevinbenton | amotoki: this will need to be back-ported into icehouse if it doesn’t make the cut | 07:34 |
| *** svilgelm has joined #openstack-neutron | 07:35 | |
| *** tomoe_ has joined #openstack-neutron | 07:35 | |
| *** marun is now known as maru_afk | 07:36 | |
| amotoki | kevinbenton: i understand using synchronized works for a workaround but I wonder something might be wrong in the design level. | 07:36 |
| *** jgallard has joined #openstack-neutron | 07:37 | |
| *** ramishra has quit IRC | 07:38 | |
| kevinbenton | amotoki: yes. there are fundamental issues with yields happening inside of transactions causing the eventlet/mysql deadlock | 07:41 |
| kevinbenton | amotoki: this plagued the ML2 driver as well | 07:42 |
| kevinbenton | amotoki: marun is looking into it now i think, but there isn’t a particularily good fix | 07:43 |
| amotoki | kevinbenton: yes. this is very tricky area. let me clarfiy one point: does eventlet/mysql deadlock occurs inside one process? | 07:44 |
| kevinbenton | that’s the only way it occurs. it happens inside of a single eventlet pool | 07:45 |
| kevinbenton | amotoki: https://docs.google.com/drawings/d/13A2x4AWbf8zmzeGApUmYVlBrW8CMTPFTCBGSP_nTzDA/edit?usp=sharing | 07:45 |
| *** tomoe_ has quit IRC | 07:47 | |
| amotoki | kevinbenton: thanks for the good pointer. that is what I expecetd. Attempting a same resource in multiple coroutine can hit the issue. | 07:47 |
| *** tomoe_ has joined #openstack-neutron | 07:47 | |
| amotoki | it really helps me understand the detail. | 07:48 |
| *** Longgeek has quit IRC | 07:48 | |
| amotoki | on the other hand, remote REST calls are placed in a transaction in most plugins/drivers. It can increase the possibility of deadlocks. | 07:49 |
| kevinbenton | amotoki: yeah, it’s unfortunate because doing the rest call inside the transaction is really nice when it can handle the failure rollback automatically | 07:50 |
| *** yfried__ has quit IRC | 07:52 | |
| amotoki | I hit the same issue in nec plugin CI before we move delete_port outside of transaction, but after fixing it I don't see any so far :-) | 07:52 |
| *** tomoe_ has quit IRC | 07:52 | |
| *** yfried__ has joined #openstack-neutron | 07:52 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: WIP Implement test https://review.openstack.org/76520 | 07:54 |
| *** tomoe_ has joined #openstack-neutron | 07:55 | |
| amotoki | kevinbenton: anyway thanks for the clarfication. your patch makes sense as a workaround. | 07:55 |
| kevinbenton | amotoki: they may be lurking there… ;-) | 07:56 |
| kevinbenton | amotoki: are the port creations asynchronous in NEC? | 07:56 |
| amotoki | kevinbenton: at now port creation is synchronous but we see race condition between dhcp-release and delete-network. | 07:57 |
| amotoki | it is similar issue because port status is updated after each operation. | 07:57 |
| *** jpich has joined #openstack-neutron | 08:02 | |
| *** Longgeek has joined #openstack-neutron | 08:02 | |
| *** ygbo has joined #openstack-neutron | 08:03 | |
| *** matsuhashi has quit IRC | 08:03 | |
| *** sridhar is now known as SridharG | 08:04 | |
| *** matsuhas_ has joined #openstack-neutron | 08:05 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Stop using portbindings_db in BSN ML2 driver https://review.openstack.org/84377 | 08:08 |
| kevinbenton | amotoki: ah, i see. another approach they were talking about was auto-retrying db operations that have a lock error, which should work as a patch as well | 08:09 |
| kevinbenton | amotoki: there would still be an occasional long delay when one does occur, but at least it won’t fail | 08:09 |
| kevinbenton | amotoki: it seems nova does this already | 08:10 |
| kevinbenton | amotoki: because mysql even warns that under high loads, it will even have it’s own internal deadlocks on indexes | 08:10 |
| kevinbenton | amotoki: so they need to be handled somehow | 08:11 |
| *** jlibosva has joined #openstack-neutron | 08:13 | |
| *** jlibosva has quit IRC | 08:19 | |
| openstackgerrit | Cedric Brandily proposed a change to openstack/neutron: Use patch ports to interconnect integration/physical bridges https://review.openstack.org/76685 | 08:19 |
| *** WackoRobie has joined #openstack-neutron | 08:20 | |
| *** jroovers has joined #openstack-neutron | 08:23 | |
| *** jroovers|afk has joined #openstack-neutron | 08:23 | |
| *** amuller has joined #openstack-neutron | 08:24 | |
| *** WackoRobie has quit IRC | 08:25 | |
| *** jroovers has quit IRC | 08:27 | |
| *** foexle has joined #openstack-neutron | 08:32 | |
| *** matsuhas_ has quit IRC | 08:33 | |
| foexle | morning guys, is there any solution to solve the SPOF of LBaaS atm ? I mean maybe in combination with pacemaker but i think if i replicate each confiuration file from host-a to host-b and switch a failover ip i dont have the corect devices although i switch the l3 routers to the correct agent right ? | 08:35 |
| *** jlibosva has joined #openstack-neutron | 08:35 | |
| foexle | so do anyone find a solution for this spof ? | 08:35 |
| *** ramishra has joined #openstack-neutron | 08:38 | |
| *** matsuhashi has joined #openstack-neutron | 08:38 | |
| *** ramishra has quit IRC | 08:43 | |
| *** saju_m has quit IRC | 08:43 | |
| *** tomoe_ has quit IRC | 08:43 | |
| *** tomoe_ has joined #openstack-neutron | 08:44 | |
| *** sandr8 has quit IRC | 08:48 | |
| *** tomoe_ has quit IRC | 08:48 | |
| *** jp_at_hp has joined #openstack-neutron | 08:53 | |
| *** jpich has quit IRC | 08:55 | |
| *** saju_m has joined #openstack-neutron | 08:56 | |
| *** afazekas has joined #openstack-neutron | 08:57 | |
| *** yonglihe_ has quit IRC | 09:00 | |
| *** overlayer has joined #openstack-neutron | 09:07 | |
| *** jpich has joined #openstack-neutron | 09:07 | |
| *** bvandenh has quit IRC | 09:10 | |
| *** lollipop has quit IRC | 09:11 | |
| *** flwang has quit IRC | 09:12 | |
| *** ramishra has joined #openstack-neutron | 09:13 | |
| *** jlibosva1 has joined #openstack-neutron | 09:14 | |
| *** svilgelm has quit IRC | 09:14 | |
| *** bashok has joined #openstack-neutron | 09:16 | |
| *** ramishra has quit IRC | 09:17 | |
| *** Moss has joined #openstack-neutron | 09:18 | |
| *** jlibosva1 has quit IRC | 09:18 | |
| *** pradipta is now known as pradipta_away | 09:19 | |
| Moss | hello, i'm looking for help in below case | 09:20 |
| Moss | https://ask.openstack.org/en/question/26096/how-to-expose-an-address-not-bound-to-a-neutron-port/ | 09:20 |
| Moss | thanks in advance for any ideas | 09:20 |
| *** WackoRobie has joined #openstack-neutron | 09:21 | |
| *** Longgeek has quit IRC | 09:22 | |
| *** lollipop has joined #openstack-neutron | 09:22 | |
| *** ramishra has joined #openstack-neutron | 09:24 | |
| *** WackoRobie has quit IRC | 09:25 | |
| *** chandankumar_ has joined #openstack-neutron | 09:25 | |
| kevinbenton | Moss: sorry i have to go shortly, but is haproxy from the loadbalancer as a service build into neutron? | 09:27 |
| *** xuhanp has quit IRC | 09:27 | |
| Moss | no, it's a separate daemon installed on lb1 and lb2 instances | 09:27 |
| kevinbenton | Moss: ok, so neutron isn’t aware of the vip at all then, right? | 09:28 |
| Moss | yes | 09:28 |
| kevinbenton | Moss: and what do you mean by “expose public virtual IP”? like output it from the heat template once the stack is complete? | 09:29 |
| Moss | yes, to be able to reach it from outside world | 09:30 |
| kevinbenton | oh wait, these instances aren’t attached directly to the 192.168.0.0/24 network are they? if those are floating IPs then they are going through a router that is reponsible for the floating IPs... | 09:32 |
| *** Longgeek has joined #openstack-neutron | 09:33 | |
| Moss | no, they are not - i'm using floating ips | 09:35 |
| *** jlibosva1 has joined #openstack-neutron | 09:36 | |
| kevinbenton | Moss: okay, then this won’t work correctly because the VIP is in the floating IP subnet but it is being realized on the inside network of the router | 09:36 |
| kevinbenton | Moss: one approach would be to make the public net a shared net as well and attach both load balancers with another interface directly to the public net | 09:37 |
| kevinbenton | so the load balancers wouldn’t be using floating IPs, they would be directly attached to that network | 09:38 |
| kevinbenton | then when they generate the VIP, it will be on the right network | 09:38 |
| *** yfujioka has joined #openstack-neutron | 09:38 | |
| Moss | yes, it should be enough for the case | 09:39 |
| Moss | thanks Kevin | 09:39 |
| *** jistr has quit IRC | 09:39 | |
| *** amuller has quit IRC | 09:41 | |
| *** jlibosva1 has quit IRC | 09:41 | |
| kevinbenton | Moss: here is a crude diagram to help visualize the floating IP topo in neutron | 09:43 |
| kevinbenton | Moss: https://docs.google.com/drawings/d/1j6T9N6_IU_QKgtOXNaO2LP4wgCHg5ccmGV_KV_dROSw/edit?usp=sharing | 09:43 |
| kevinbenton | Moss: you may run into some issues with security groups when the load balancers pretend to have another IP that wasn’t assigned to them by neutron | 09:46 |
| *** morganfainberg is now known as morganfainberg_Z | 09:46 | |
| kevinbenton | Moss: so make them very permissive or even disable them until you get it working and then re-enable them | 09:46 |
| kevinbenton | Moss: ok, that should give you a good start. I have to go now. ttyl | 09:47 |
| Moss | kevinbenton: yes, i need to think it thru - thanks a lot! | 09:47 |
| kevinbenton | Moss: i just updated to directly reflect what you will probably need to do | 09:52 |
| *** sphoorti has joined #openstack-neutron | 09:53 | |
| kevinbenton | Moss: either that or use the neutron load balancer as a service | 09:53 |
| kevinbenton | Moss: ok, have to leave for real now :-) | 09:53 |
| Moss | kevinbenton: all is clear now :) i have work to do Bye | 09:53 |
| *** chandankumar_ has quit IRC | 09:54 | |
| *** Jabadia_ has joined #openstack-neutron | 09:55 | |
| *** Jabadia has quit IRC | 09:55 | |
| *** Moss has quit IRC | 09:56 | |
| *** jistr has joined #openstack-neutron | 09:59 | |
| *** ihrachys is now known as ihrachys|lunch | 09:59 | |
| *** Jabadia_ has quit IRC | 10:02 | |
| openstackgerrit | A change was merged to openstack/neutron: BigSwitch: Semaphore on port status update https://review.openstack.org/84368 | 10:07 |
| *** moss has joined #openstack-neutron | 10:09 | |
| *** safchain has joined #openstack-neutron | 10:10 | |
| *** matsuhashi has quit IRC | 10:14 | |
| *** amuller has joined #openstack-neutron | 10:17 | |
| *** jgallard has quit IRC | 10:20 | |
| *** WackoRobie has joined #openstack-neutron | 10:21 | |
| *** pcm_ has joined #openstack-neutron | 10:24 | |
| *** yamahata has quit IRC | 10:24 | |
| *** pcm_ has quit IRC | 10:24 | |
| *** pcm_ has joined #openstack-neutron | 10:24 | |
| *** WackoRobie has quit IRC | 10:26 | |
| *** dave_tucker_zzz is now known as dave_tucker | 10:26 | |
| *** overlayer has quit IRC | 10:26 | |
| *** roaet has quit IRC | 10:29 | |
| *** overlayer has joined #openstack-neutron | 10:31 | |
| *** leseb has quit IRC | 10:32 | |
| *** leseb has joined #openstack-neutron | 10:32 | |
| *** gdubreui has joined #openstack-neutron | 10:34 | |
| *** overlayer has quit IRC | 10:34 | |
| *** gdubreui has quit IRC | 10:35 | |
| *** jlibosva1 has joined #openstack-neutron | 10:36 | |
| *** leseb has quit IRC | 10:37 | |
| *** evgenyf has quit IRC | 10:37 | |
| *** overlayer has joined #openstack-neutron | 10:40 | |
| *** jlibosva1 has quit IRC | 10:41 | |
| *** overlayer has quit IRC | 10:42 | |
| *** roaet has joined #openstack-neutron | 10:43 | |
| *** evgenyf has joined #openstack-neutron | 10:44 | |
| *** overlayer has joined #openstack-neutron | 10:45 | |
| *** Jabadia has joined #openstack-neutron | 10:45 | |
| openstackgerrit | Yuriy Taraday proposed a change to openstack/neutron: Abstract out root_helper calls to classes https://review.openstack.org/82787 | 10:48 |
| *** yfujioka has quit IRC | 10:53 | |
| *** overlayer has quit IRC | 10:59 | |
| sphoorti | salv-orlando: Hello! I am an OPW applicant and as a part of my application I had written a unit test for the add_vxlan function in test_linux_ip_lib. Could you please review it when you have time ? https://review.openstack.org/#/c/80554/ | 11:01 |
| *** leseb has joined #openstack-neutron | 11:01 | |
| *** Jabadia has quit IRC | 11:01 | |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: L2 pop removes unuseful regular mac learning rules https://review.openstack.org/83053 | 11:05 |
| *** leseb has quit IRC | 11:05 | |
| *** roeyc has joined #openstack-neutron | 11:06 | |
| *** baoli has quit IRC | 11:15 | |
| sphoorti | salv-orlando: thank you :) | 11:15 |
| irenab | salv-orlando: hi | 11:20 |
| *** WackoRobie has joined #openstack-neutron | 11:22 | |
| *** ihrachys|lunch is now known as ihrachys | 11:24 | |
| *** jgallard has joined #openstack-neutron | 11:24 | |
| *** WackoRobie has quit IRC | 11:27 | |
| *** alexpilotti has joined #openstack-neutron | 11:34 | |
| *** jlibosva1 has joined #openstack-neutron | 11:37 | |
| *** xianghui has quit IRC | 11:41 | |
| *** jlibosva1 has quit IRC | 11:42 | |
| *** saju_m has quit IRC | 11:44 | |
| openstackgerrit | A change was merged to openstack/neutron: Invoke _process_l3_create within plugin session https://review.openstack.org/84038 | 11:47 |
| openstackgerrit | A change was merged to openstack/neutron: Open Juno development https://review.openstack.org/83434 | 11:47 |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Updated from global requirements https://review.openstack.org/83474 | 11:51 |
| *** jgallard has quit IRC | 11:52 | |
| *** jgallard has joined #openstack-neutron | 11:52 | |
| *** linuxaddicts has quit IRC | 11:53 | |
| *** ramishra has quit IRC | 11:56 | |
| *** saju_m has joined #openstack-neutron | 11:56 | |
| *** leseb has joined #openstack-neutron | 11:57 | |
| *** mflobo has joined #openstack-neutron | 11:57 | |
| *** baoli has joined #openstack-neutron | 11:59 | |
| *** saju_m has quit IRC | 12:02 | |
| *** afazekas has quit IRC | 12:04 | |
| moss | kevinbenton: following your advice i changed the case and used built-in lbaas (http://openstack.redhat.com/LBaaS) - thanks for clearing this out for me Kevin | 12:05 |
| *** linuxaddicts has joined #openstack-neutron | 12:08 | |
| *** overlayer has joined #openstack-neutron | 12:10 | |
| *** leseb has quit IRC | 12:10 | |
| *** amotoki has quit IRC | 12:11 | |
| *** leseb has joined #openstack-neutron | 12:11 | |
| *** leseb has quit IRC | 12:16 | |
| *** afazekas has joined #openstack-neutron | 12:18 | |
| *** overlayer has quit IRC | 12:18 | |
| *** saju_m has joined #openstack-neutron | 12:21 | |
| openstackgerrit | A change was merged to openstack/neutron: Add unit test for add_vxlan in test_linux_ip_lib https://review.openstack.org/80554 | 12:21 |
| *** dims has quit IRC | 12:22 | |
| *** saju_m has quit IRC | 12:23 | |
| *** WackoRobie has joined #openstack-neutron | 12:23 | |
| *** WackoRob_ has joined #openstack-neutron | 12:24 | |
| *** leseb has joined #openstack-neutron | 12:26 | |
| *** markmcclain has joined #openstack-neutron | 12:28 | |
| *** WackoRobie has quit IRC | 12:28 | |
| openstackgerrit | Marios Andreou proposed a change to openstack/neutron: Validate CIDR given as ip-prefix in security-group-rule-create https://review.openstack.org/59212 | 12:28 |
| *** dims_ has joined #openstack-neutron | 12:34 | |
| *** tomoe_ has joined #openstack-neutron | 12:36 | |
| *** ramishra has joined #openstack-neutron | 12:37 | |
| *** jlibosva1 has joined #openstack-neutron | 12:38 | |
| *** b3nt_pin has joined #openstack-neutron | 12:38 | |
| *** saju_m has joined #openstack-neutron | 12:39 | |
| *** b3nt_pin is now known as beagles | 12:39 | |
| *** jlibosva1 has quit IRC | 12:42 | |
| *** jgrimm has joined #openstack-neutron | 12:43 | |
| *** jprovazn has joined #openstack-neutron | 12:45 | |
| *** overlayer has joined #openstack-neutron | 12:47 | |
| anteaya | congratulations sphoorti | 12:49 |
| sphoorti | thank you anteaya :) | 12:50 |
| sphoorti | it feels nice | 12:50 |
| sphoorti | my first patch | 12:50 |
| anteaya | I remember that first patch feeling | 12:50 |
| anteaya | you worked hard | 12:50 |
| anteaya | well done | 12:50 |
| *** thuc has joined #openstack-neutron | 12:50 | |
| *** jrbudnack has joined #openstack-neutron | 12:52 | |
| sphoorti | thank you so much anteaya :) | 12:52 |
| *** thuc_ has joined #openstack-neutron | 12:52 | |
| anteaya | :D | 12:52 |
| sphoorti | anteaya: I also wrote a blogpost about setting up neutron in devstack http://sphoorti.github.io/blog/devstack/devstack_setup.html | 12:54 |
| *** overlayer has quit IRC | 12:54 | |
| *** thuc has quit IRC | 12:55 | |
| anteaya | nice | 12:57 |
| sphoorti | thank you | 12:57 |
| anteaya | I have a request | 12:57 |
| anteaya | do you mind using git://git.openstack.org/openstack-dev/devstack as you git clone url? | 12:58 |
| anteaya | since openstack is opensource we prefer to use and promote the use of opensource tools | 12:58 |
| anteaya | git.openstack.org or git.o.o is our internal git server for all our repos | 12:58 |
| anteaya | and we do prefer that it be referenced in documents if this nature | 12:59 |
| anteaya | if you dont' mind using it | 12:59 |
| anteaya | it is a cgit server | 12:59 |
| anteaya | well actually 5 of them | 12:59 |
| anteaya | I stand corrected 6 of them: http://cacti.openstack.org/cacti/graph_view.php?action=tree&tree_id=2 | 13:00 |
| sphoorti | anteaya: I had referred this http://devstack.org/ | 13:00 |
| *** itzikb has joined #openstack-neutron | 13:01 | |
| *** heyongli has joined #openstack-neutron | 13:01 | |
| sphoorti | I have never tried with git://git.openstack.org/openstack-dev/devstack | 13:01 |
| anteaya | okay | 13:01 |
| anteaya | also you might be interested in looking at http://git.openstack.org/cgit/openstack/oslo-incubator/tree/TESTING.rst | 13:02 |
| anteaya | it has some commands for running an individual test with tox | 13:02 |
| *** sballe has joined #openstack-neutron | 13:02 | |
| sphoorti | git.openstack is the internal server is it? Then why isnt it referenced in the official docs anteaya ? | 13:02 |
| anteaya | generally programs are moving away from run_tests.sh | 13:02 |
| anteaya | sphoorti: whick official docs? | 13:02 |
| anteaya | which | 13:02 |
| *** jroovers|afk has quit IRC | 13:03 | |
| sphoorti | the devstack guide? | 13:03 |
| anteaya | probably because noone has asked dean troyer about it | 13:03 |
| anteaya | join me in #openstack-qa and we can do so | 13:03 |
| *** tomoe_ has quit IRC | 13:04 | |
| *** tomoe_ has joined #openstack-neutron | 13:05 | |
| *** rand738 has quit IRC | 13:08 | |
| *** rand738 has joined #openstack-neutron | 13:08 | |
| anteaya | sphoorti: I have no idea what timezone he works in | 13:08 |
| djoreilly | qiuyu: the allowed address pairs extension seems to allow non-admin tenants to override the anti-ip-spoof rules? | 13:09 |
| anteaya | it might take several hours for a response, stay in the channel or read the logs: http://eavesdrop.openstack.org/irclogs/%23openstack-qa/ | 13:09 |
| *** jecarey__ has quit IRC | 13:09 | |
| anteaya | then we will both know | 13:09 |
| *** tomoe_ has quit IRC | 13:10 | |
| sphoorti | anteaya: I think he must be in PCT | 13:10 |
| sphoorti | *Pacific time zone | 13:10 |
| sphoorti | I shall stay on the channel | 13:11 |
| *** WackoRob_ has quit IRC | 13:11 | |
| *** WackoRobie has joined #openstack-neutron | 13:12 | |
| *** afazekas has quit IRC | 13:13 | |
| *** saju_m has quit IRC | 13:15 | |
| *** mflobo has quit IRC | 13:16 | |
| *** WackoRobie has quit IRC | 13:16 | |
| *** ram5391 has joined #openstack-neutron | 13:20 | |
| *** saju_m has joined #openstack-neutron | 13:28 | |
| *** xuhanp has joined #openstack-neutron | 13:30 | |
| *** moss has left #openstack-neutron | 13:34 | |
| *** jgrimm has quit IRC | 13:34 | |
| *** changbl has quit IRC | 13:35 | |
| *** tomoe_ has joined #openstack-neutron | 13:35 | |
| *** jorgem has joined #openstack-neutron | 13:37 | |
| *** julim has joined #openstack-neutron | 13:38 | |
| *** jorgem1 has joined #openstack-neutron | 13:38 | |
| *** jlibosva1 has joined #openstack-neutron | 13:39 | |
| *** mwagner_zzz has quit IRC | 13:40 | |
| *** thuc_ has quit IRC | 13:40 | |
| *** thuc has joined #openstack-neutron | 13:41 | |
| *** jorgem has quit IRC | 13:42 | |
| *** sballe has quit IRC | 13:43 | |
| *** jlibosva1 has quit IRC | 13:44 | |
| *** nplanel_ has joined #openstack-neutron | 13:45 | |
| *** thuc has quit IRC | 13:45 | |
| *** markmcclain has quit IRC | 13:45 | |
| *** thuc has joined #openstack-neutron | 13:45 | |
| *** afazekas has joined #openstack-neutron | 13:46 | |
| *** banix has joined #openstack-neutron | 13:52 | |
| *** jecarey has joined #openstack-neutron | 13:55 | |
| *** jroovers has joined #openstack-neutron | 13:55 | |
| *** jroovers|afk has joined #openstack-neutron | 13:56 | |
| *** afazekas has quit IRC | 13:57 | |
| openstackgerrit | Simon Pasquier proposed a change to openstack/neutron: Fix function parsing the kernel version https://review.openstack.org/84441 | 13:57 |
| *** WackoRobie has joined #openstack-neutron | 13:58 | |
| *** jroovers has quit IRC | 14:00 | |
| openstackgerrit | A change was merged to openstack/neutron: Updated from global requirements https://review.openstack.org/83474 | 14:00 |
| *** jecarey has quit IRC | 14:02 | |
| *** jecarey has joined #openstack-neutron | 14:03 | |
| openstackgerrit | Simon Pasquier proposed a change to openstack/neutron: Fix function parsing the kernel version https://review.openstack.org/84441 | 14:03 |
| *** heyongli has quit IRC | 14:05 | |
| *** aveiga has joined #openstack-neutron | 14:05 | |
| *** tomoe_ has quit IRC | 14:06 | |
| *** tomoe_ has joined #openstack-neutron | 14:07 | |
| *** lollipop has quit IRC | 14:08 | |
| *** flwang has joined #openstack-neutron | 14:10 | |
| *** tomoe_ has quit IRC | 14:11 | |
| *** tomoe_ has joined #openstack-neutron | 14:11 | |
| *** saju_m has quit IRC | 14:14 | |
| *** irenab has quit IRC | 14:15 | |
| *** jobewan has joined #openstack-neutron | 14:15 | |
| *** alagalah has joined #openstack-neutron | 14:17 | |
| *** otherwiseguy has joined #openstack-neutron | 14:17 | |
| *** roeyc has quit IRC | 14:18 | |
| *** jgrimm has joined #openstack-neutron | 14:20 | |
| *** markmcclain has joined #openstack-neutron | 14:21 | |
| *** zhipeng has joined #openstack-neutron | 14:23 | |
| *** alagalah has quit IRC | 14:28 | |
| *** skraynev is now known as skraynev_afk | 14:30 | |
| *** manishg has joined #openstack-neutron | 14:30 | |
| *** sballe has joined #openstack-neutron | 14:31 | |
| *** manishg_ has joined #openstack-neutron | 14:34 | |
| *** manishg has quit IRC | 14:34 | |
| *** jprovazn is now known as jprovazn_afk | 14:35 | |
| *** afazekas has joined #openstack-neutron | 14:37 | |
| *** jlibosva1 has joined #openstack-neutron | 14:40 | |
| *** manishg_ has quit IRC | 14:41 | |
| *** thuc_ has joined #openstack-neutron | 14:42 | |
| openstackgerrit | Mohammad Banikazemi proposed a change to openstack/neutron: Removing vim header lines https://review.openstack.org/84186 | 14:43 |
| *** jlibosva1 has quit IRC | 14:44 | |
| *** thuc has quit IRC | 14:45 | |
| *** thuc_ has quit IRC | 14:46 | |
| openstackgerrit | Jakub Libosvar proposed a change to openstack/neutron: Use os.uname() instead of calling uname in subprocess https://review.openstack.org/84462 | 14:47 |
| xuhanp | sc68cal, ping | 14:48 |
| sc68cal | pong | 14:48 |
| xuhanp | sc68cal, will be great if you can have a look at Akihiro's comment in patch 4 of https://review.openstack.org/#/c/75871/ | 14:48 |
| sc68cal | sure - assertRaises usage? | 14:49 |
| xuhanp | my trouble is with the specific exception | 14:49 |
| sc68cal | meaning which one to use? | 14:51 |
| xuhanp | During debug, I found out that the exception is not related to the test case (which makes sense since I didn't mock anything) | 14:51 |
| xuhanp | so I am not sure if I should mock the exception from server side | 14:52 |
| xuhanp | I didn't find something similar in the client test code. | 14:52 |
| sc68cal | what is the type of the exception that was thrown | 14:53 |
| xuhanp | I was trying to find that. One sec | 14:53 |
| *** mlavalle has joined #openstack-neutron | 14:54 | |
| openstackgerrit | Claudiu Belu proposed a change to openstack/neutron: Fixes Hyper-V agent security group ICMP rules https://review.openstack.org/84464 | 14:54 |
| *** manishg has joined #openstack-neutron | 14:55 | |
| xuhanp | sc68cal, sorry I could not find it in the code because the log is on my other computer. But something similar to Connection failed | 14:56 |
| *** lollipop has joined #openstack-neutron | 14:56 | |
| *** thedodd has joined #openstack-neutron | 14:56 | |
| sc68cal | xuhanp: OK - let me pull down the code and tinker with it this week | 14:58 |
| xuhanp | sc68cal, I got it now. it's ConnectionFailed from client.py | 14:58 |
| xuhanp | raise exceptions.ConnectionFailed(reason=e) | 14:58 |
| xuhanp | sc68cal, thanks a lot! | 14:58 |
| *** thuc has joined #openstack-neutron | 15:00 | |
| *** sphoorti_ has joined #openstack-neutron | 15:00 | |
| *** manishg has quit IRC | 15:00 | |
| *** sphoorti has quit IRC | 15:00 | |
| *** jroovers|afk is now known as jroovers | 15:00 | |
| *** thuc has quit IRC | 15:01 | |
| *** thuc has joined #openstack-neutron | 15:02 | |
| *** sbfox has joined #openstack-neutron | 15:02 | |
| *** carl_baldwin has joined #openstack-neutron | 15:03 | |
| *** djoreilly has quit IRC | 15:04 | |
| openstackgerrit | Armando Migliaccio proposed a change to openstack/neutron: NSX: Fix pagination support https://review.openstack.org/82218 | 15:05 |
| *** geekinutah has joined #openstack-neutron | 15:08 | |
| *** sphoorti__ has joined #openstack-neutron | 15:11 | |
| *** sphoorti_ has quit IRC | 15:12 | |
| *** chandankumar_ has joined #openstack-neutron | 15:12 | |
| *** ramishra has quit IRC | 15:16 | |
| *** hrushig has joined #openstack-neutron | 15:17 | |
| *** prad has joined #openstack-neutron | 15:17 | |
| *** armax has joined #openstack-neutron | 15:17 | |
| *** RockKuo has joined #openstack-neutron | 15:18 | |
| *** aveiga has quit IRC | 15:21 | |
| *** zhipeng has quit IRC | 15:22 | |
| *** alexpilotti has quit IRC | 15:22 | |
| *** nplanel_ has quit IRC | 15:23 | |
| *** nplanel_ has joined #openstack-neutron | 15:24 | |
| *** markwash has quit IRC | 15:25 | |
| *** markwash has joined #openstack-neutron | 15:26 | |
| openstackgerrit | A change was merged to openstack/neutron: Add missing DB migrations for BSN ML2 plugin https://review.openstack.org/84313 | 15:34 |
| *** djoreilly has joined #openstack-neutron | 15:35 | |
| *** itzikb has quit IRC | 15:38 | |
| *** xuhanp has quit IRC | 15:38 | |
| *** networkstatic has joined #openstack-neutron | 15:39 | |
| *** jlibosva1 has joined #openstack-neutron | 15:40 | |
| *** lollipop has quit IRC | 15:41 | |
| *** networkstatic has quit IRC | 15:43 | |
| *** alagalah has joined #openstack-neutron | 15:44 | |
| *** jlibosva1 has quit IRC | 15:45 | |
| *** tvardeman has joined #openstack-neutron | 15:46 | |
| *** thuc_ has joined #openstack-neutron | 15:47 | |
| *** blogan has joined #openstack-neutron | 15:49 | |
| *** thuc has quit IRC | 15:50 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Stop using portbindings_db in BSN ML2 driver https://review.openstack.org/84377 | 15:50 |
| *** leseb has quit IRC | 15:51 | |
| *** thuc_ has quit IRC | 15:51 | |
| *** leseb has joined #openstack-neutron | 15:51 | |
| *** casmith32 has joined #openstack-neutron | 15:54 | |
| *** bashok has quit IRC | 15:55 | |
| *** otherwiseguy has quit IRC | 15:55 | |
| *** SridharG has quit IRC | 15:55 | |
| *** _cjones_ has joined #openstack-neutron | 15:56 | |
| *** jlibosva has quit IRC | 15:57 | |
| *** changbl has joined #openstack-neutron | 15:59 | |
| *** blogan has quit IRC | 15:59 | |
| *** bashok has joined #openstack-neutron | 16:00 | |
| *** jgallard has quit IRC | 16:00 | |
| *** safchain has quit IRC | 16:04 | |
| *** sballe_ has joined #openstack-neutron | 16:05 | |
| *** sballe has quit IRC | 16:05 | |
| *** jlibosva has joined #openstack-neutron | 16:05 | |
| *** sbfox has quit IRC | 16:09 | |
| *** amotoki has joined #openstack-neutron | 16:10 | |
| *** Sukhdev has joined #openstack-neutron | 16:10 | |
| Sukhdev | rkukura: Ping | 16:11 |
| rkukura | Sukhdev: pong | 16:13 |
| *** sbfox has joined #openstack-neutron | 16:13 | |
| *** mwagner_zzz has joined #openstack-neutron | 16:16 | |
| *** jistr has quit IRC | 16:17 | |
| *** RockKuo has quit IRC | 16:17 | |
| *** thuc has joined #openstack-neutron | 16:18 | |
| *** thuc has quit IRC | 16:18 | |
| *** bashok has quit IRC | 16:18 | |
| *** thuc has joined #openstack-neutron | 16:19 | |
| *** devvesa has quit IRC | 16:20 | |
| openstackgerrit | A change was merged to openstack/neutron: Fix function parsing the kernel version https://review.openstack.org/84441 | 16:27 |
| *** jroovers has quit IRC | 16:29 | |
| *** sbfox1 has joined #openstack-neutron | 16:30 | |
| yfried__ | mlavalle: ping | 16:30 |
| *** bjornar has joined #openstack-neutron | 16:32 | |
| *** maru_afk is now known as marun | 16:32 | |
| marun | carl_baldwin: ping | 16:33 |
| carl_baldwin | marun: pong | 16:33 |
| *** sbfox has quit IRC | 16:33 | |
| openstackgerrit | A change was merged to openstack/neutron: Move test_ovs_lib to tests/unit/agent/linux https://review.openstack.org/83896 | 16:33 |
| *** Sukhdev has quit IRC | 16:33 | |
| marun | carl_baldwin: I heard through the grapevine that multiple api+rpc workers is breaking delivery of notification to nova. | 16:34 |
| marun | carl_baldwin: I think the first step in resolving that is going to be creating a check job that will turn on multiple workers and we can use that job as a way of validating the eventual fix. | 16:34 |
| marun | carl_baldwin: I can take responsibility for taht. | 16:34 |
| *** sballe_ has quit IRC | 16:34 | |
| marun | carl_baldwin: is this on your radar, or is someone else working on it? | 16:34 |
| carl_baldwin | marun: That was not on my radar. I did have a check job somewhere. | 16:35 |
| marun | carl_baldwin: ah, so much the better | 16:35 |
| marun | carl_baldwin: I think if we can get the job running we can advocate for changing the defaults to at least 1 for both api and rpc | 16:35 |
| marun | carl_baldwin: it's too important not to have it tested in the gate | 16:36 |
| carl_baldwin | marun: https://review.openstack.org/#/c/76021 | 16:36 |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Split API and RPC workers in to one separate process each (WIP) https://review.openstack.org/76021 | 16:37 |
| carl_baldwin | I just restored and rebased it. | 16:37 |
| marun | carl_baldwin: ah, ok. I was thinking we want a separate check job so that it can be run against every patch for a while. | 16:37 |
| marun | carl_baldwin: thoughts? | 16:37 |
| *** dave_tucker is now known as dave_tucker_zzz | 16:37 | |
| carl_baldwin | Sure. | 16:37 |
| *** manishg has joined #openstack-neutron | 16:38 | |
| marun | carl_baldwin: I guess the patch in question should probably be stabilized first. | 16:38 |
| carl_baldwin | marun: That is kind of what I was just thinking. | 16:38 |
| marun | carl_baldwin: do you know anything about how eventlet threads respond to forks? | 16:39 |
| marun | carl_baldwin: Apparently the thread that delivers notifications to nova is started in the parent but does not appear to run in the children. | 16:39 |
| carl_baldwin | marun: I had some thoughts on that at the time. Let me look at my notes. | 16:39 |
| marun | carl_baldwin: great! | 16:40 |
| carl_baldwin | marun: I think you're right on about event let not running anymore in the children. I have a vague recollection about hitting that problem. | 16:41 |
| *** leseb has quit IRC | 16:41 | |
| *** leseb has joined #openstack-neutron | 16:41 | |
| *** jlibosva1 has joined #openstack-neutron | 16:42 | |
| carl_baldwin | marun: It was the motivation behind add "start_rpc_listener" and calling it after the fork. | 16:45 |
| *** singhs has joined #openstack-neutron | 16:45 | |
| *** leseb has quit IRC | 16:45 | |
| carl_baldwin | marun: Do you happen to know where the the nova notification thread is started off the top of your head? | 16:46 |
| *** jlibosva1 has quit IRC | 16:46 | |
| carl_baldwin | marun: nm. I've found found what I'm looking for. | 16:47 |
| marun | carl_baldwin: ok | 16:47 |
| *** otherwiseguy has joined #openstack-neutron | 16:50 | |
| *** jrbudnack has quit IRC | 16:50 | |
| *** singhs has quit IRC | 16:54 | |
| *** dvorkinista has joined #openstack-neutron | 16:56 | |
| *** singhs has joined #openstack-neutron | 16:57 | |
| *** Longgeek has quit IRC | 16:59 | |
| *** Longgeek has joined #openstack-neutron | 16:59 | |
| *** SumitNaiksatam has quit IRC | 17:00 | |
| *** Longgeek has quit IRC | 17:01 | |
| *** arosen has quit IRC | 17:02 | |
| *** armax has quit IRC | 17:02 | |
| *** rms_13 has joined #openstack-neutron | 17:03 | |
| carl_baldwin | marun: Are you familiar with how sqlalchemy.event works? | 17:03 |
| marun | carl_baldwin: roughly | 17:03 |
| *** armax has joined #openstack-neutron | 17:04 | |
| *** afazekas is now known as afazekas_noteboo | 17:05 | |
| *** amotoki has quit IRC | 17:05 | |
| *** Sukhdev has joined #openstack-neutron | 17:05 | |
| *** arosen has joined #openstack-neutron | 17:05 | |
| carl_baldwin | marun: I guess my question is whether you know if the events are being batched up in nova.Notifier and not sent or if they're not being batched at all. | 17:06 |
| marun | carl_baldwin: events are being batched | 17:06 |
| *** ygbo has quit IRC | 17:06 | |
| marun | carl_baldwin: the intent was to avoid network io in db transactions | 17:06 |
| carl_baldwin | marun: Yeah, I just wanted to rule out any potential problem with the sqlalchemy event listener that is responsible for batching the events in NeutronDbPluginV2. | 17:07 |
| marun | carl_baldwin: oh, you mean sqlalchemy batching? | 17:08 |
| marun | carl_baldwin: I'm pretty sure that is synchronous | 17:08 |
| carl_baldwin | marun: I think you're right. I'm just trying to catch up and wrap my head around it all. | 17:08 |
| marun | carl_baldwin: the batching is done outside of sqlalchemy. | 17:08 |
| *** dave_tucker_zzz is now known as dave_tucker | 17:09 | |
| marun | carl_baldwin: arosen is the expert since he wrote it | 17:09 |
| *** Longgeek has joined #openstack-neutron | 17:09 | |
| arosen | carl_baldwin: its done outside of a transaction | 17:09 |
| arosen | it's all run from a seperate eventloop | 17:09 |
| *** Sukhdev has quit IRC | 17:09 | |
| carl_baldwin | arosen: Are you talking about the event loop started in nova.Notifier.__init__? | 17:10 |
| arosen | carl_baldwin: sorry what's the question i don't have the full conv just joined. | 17:10 |
| mlavalle | yfried: sorry, I just saw your ping. I have to go out now. I'll try to ping you in about 1 hour. If I don't reach you, let's try tomorrow morning. I am at UTC - 5 | 17:10 |
| arosen | carl_baldwin: events are pushed into: self.pending_events = []\ | 17:11 |
| *** dave_tucker is now known as dave_tucker_zzz | 17:11 | |
| openstackgerrit | Matthew Treinish proposed a change to openstack/neutron: Enable hacking H301 check https://review.openstack.org/84258 | 17:11 |
| arosen | then sent in the send_events() method which gets called periodically | 17:11 |
| carl_baldwin | arosen: No problem. I think the working theory is that the thread that runs send_events periodically stops when rpc workers are spawned as subprocesses when rpc_workers is set > 0. | 17:12 |
| *** sbfox1 has quit IRC | 17:12 | |
| *** nplanel_ has quit IRC | 17:12 | |
| carl_baldwin | I'm just trying to wrap my head around that theory and rule out anything else that might be happening. | 17:12 |
| *** mlavalle has quit IRC | 17:13 | |
| *** harlowja_away is now known as harlowja | 17:14 | |
| *** mwagner_zzz is now known as mwagner_ | 17:14 | |
| openstackgerrit | Aaron Rosen proposed a change to openstack/neutron: Test https://review.openstack.org/84501 | 17:15 |
| *** SumitNaiksatam has joined #openstack-neutron | 17:15 | |
| arosen | carl_baldwin: interesting so ^ might not work? | 17:15 |
| carl_baldwin | That is the idea. | 17:17 |
| *** jpich has quit IRC | 17:18 | |
| carl_baldwin | marun: arosen: I thought there was something in the ProcessLauncher code that took care of event let after the fork. A vague memory. I'm going to go look. | 17:19 |
| marun | carl_baldwin: the other alternative is to ensure that eventlet threads are only started post-fork | 17:20 |
| *** jlibosva has quit IRC | 17:20 | |
| carl_baldwin | marun: You're right. I wonder how hard that will be considering threads can be started anywhere. But, maybe it is doable if we can do the fork earlier. | 17:21 |
| *** amuller has quit IRC | 17:22 | |
| marun | carl_baldwin: We could patch eventlet.spawn to raise exceptions pre-fork... | 17:22 |
| *** Longgeek_ has joined #openstack-neutron | 17:23 | |
| *** Longgeek has quit IRC | 17:23 | |
| *** alexpilotti has joined #openstack-neutron | 17:25 | |
| carl_baldwin | Like a monkey patch or just a local patch to find all of the pre-fork spawns? | 17:25 |
| *** afazekas_noteboo is now known as _afazekas | 17:25 | |
| marun | carl_baldwin: I'm not sure there's a difference really. | 17:27 |
| marun | carl_baldwin: it would be a patch that would be removed immediately on fork | 17:28 |
| *** luqas has quit IRC | 17:28 | |
| marun | carl_baldwin: and only applied if forking is to be used | 17:28 |
| carl_baldwin | marun: I think this call is what stops event let threads: "eventlet.hubs.use_hub()" in neutron/openstack/common/service.py:497 | 17:28 |
| carl_baldwin | So, I guess the code that I thought was fixing threads could be what is actually stopping them. | 17:29 |
| *** thedodd has quit IRC | 17:29 | |
| carl_baldwin | marun: From the event let documentation: "Make sure to do this before the application starts doing any I/O! Calling use_hub completely eliminates the old hub, and any file descriptors or timers that it had been managing will be forgotten." | 17:29 |
| marun | carl_baldwin: ah, ok | 17:29 |
| carl_baldwin | I did not understand this before. | 17:30 |
| marun | carl_baldwin: eventlet is a bottomless pit of misunderstanding | 17:30 |
| carl_baldwin | haha | 17:30 |
| carl_baldwin | So, we could either go your route and make sure fork comes before any spawn. Or, we find a different way to accomplish what the openstack common code is trying to do with this call. | 17:31 |
| *** morganfainberg_Z is now known as morganfainberg | 17:31 | |
| *** sandr8 has joined #openstack-neutron | 17:31 | |
| carl_baldwin | marun: The line number up there in service.py should have been 287. | 17:32 |
| marun | carl_baldwin: I'm not sure we can avoid forking pre-spawn… Even if we could change the use_hub behaviour, how could we go back and respawn things? | 17:35 |
| carl_baldwin | marun: I don't know. | 17:37 |
| carl_baldwin | marun: I'm not sure yet how easy it will be to move spawns after the fork given the current code structure. Just trying to explore options. | 17:41 |
| *** jlibosva has joined #openstack-neutron | 17:43 | |
| carl_baldwin | marun: I think the problem boils down to where the code loads the plugin. | 17:46 |
| *** chandankumar_ has quit IRC | 17:46 | |
| marun | carl_baldwin: Why is the plugin loaded pre-fork at all? | 17:46 |
| marun | carl_baldwin: is it a matter of validating config? | 17:47 |
| marun | carl_baldwin: it may be that we have to introduce the idea of post-fork configuration where the plugins could initialize threads separately from initial init | 17:48 |
| rms_13 | obondarev: Could you please take a look at my response to your comments on https://review.openstack.org/#/c/84304/; thanks | 17:48 |
| *** jlibosva has quit IRC | 17:48 | |
| *** singhs has quit IRC | 17:51 | |
| *** singhs has joined #openstack-neutron | 17:52 | |
| *** jprovazn_afk is now known as jprovazn | 17:52 | |
| *** Longgeek has joined #openstack-neutron | 17:53 | |
| carl_baldwin | marun: My addition of "start_rpc_listener" in the plugin was to move certain things post-fork. I didn't catch the spawning of the nova notifier in that because it goes one layer deeper in the class hierarchy, the NeutronDbPluginV2 class. | 17:53 |
| *** Longgeek_ has quit IRC | 17:54 | |
| marun | carl_baldwin: I guess it makes sense to selectively move it post-fork then. | 17:54 |
| marun | carl_baldwin: the fact that this was a problem was due to a gap in testing | 17:54 |
| marun | carl_baldwin: future changes would hopefully fail those tests rather than requiring explicit patching | 17:55 |
| carl_baldwin | marun: Yes, there was a gap in testing. | 17:55 |
| carl_baldwin | marun: Would you like to work on a patch or would you like me to do it? | 17:55 |
| marun | carl_baldwin: if you have time, I think your continued maintenance of the multiple workers makes the most sense. | 17:56 |
| marun | carl_baldwin: if not, I can make sure it happens. | 17:56 |
| carl_baldwin | I'm happy to fix my own messes. :) | 17:56 |
| carl_baldwin | I'm sure I'll have something up today for review. | 17:57 |
| marun | carl_baldwin: not your mess, but thank you for taking responsibility :) | 17:57 |
| marun | carl_baldwin: I'll work on providing rapid turnaround. | 17:57 |
| carl_baldwin | marun: Glad to help. I'll tag you as a reviewer. | 17:58 |
| *** rotbeard has quit IRC | 17:58 | |
| *** sbfox has joined #openstack-neutron | 17:58 | |
| marun | carl_baldwin: great! | 17:58 |
| *** mtreinish has quit IRC | 18:00 | |
| *** mwagner_ is now known as mwagner_afk | 18:00 | |
| *** dave_tucker_zzz is now known as dave_tucker | 18:00 | |
| *** mtreinish has joined #openstack-neutron | 18:01 | |
| HenryG | arosen: salv-orlando: who can look into NSX plugin migration issues? See bug #1298456 | 18:01 |
| *** sbfox has quit IRC | 18:01 | |
| *** sridhar has joined #openstack-neutron | 18:03 | |
| *** julim has quit IRC | 18:04 | |
| *** sandr8 has quit IRC | 18:04 | |
| *** otherwiseguy has quit IRC | 18:05 | |
| *** julim has joined #openstack-neutron | 18:06 | |
| openstackgerrit | A change was merged to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/84366 | 18:13 |
| *** baoli has quit IRC | 18:13 | |
| *** Sukhdev has joined #openstack-neutron | 18:14 | |
| *** jlibosva has joined #openstack-neutron | 18:15 | |
| *** sbfox has joined #openstack-neutron | 18:22 | |
| *** sn6i23a has joined #openstack-neutron | 18:22 | |
| *** thuc has quit IRC | 18:22 | |
| *** thuc has joined #openstack-neutron | 18:24 | |
| openstackgerrit | Jakub Libosvar proposed a change to openstack/neutron: Use os.uname() instead of calling uname in subprocess https://review.openstack.org/84462 | 18:24 |
| *** thuc has quit IRC | 18:24 | |
| *** thuc has joined #openstack-neutron | 18:24 | |
| *** singhs has quit IRC | 18:32 | |
| *** sbfox has quit IRC | 18:32 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron: Remove auto-generation of db schema from models at startup https://review.openstack.org/40296 | 18:34 |
| *** Longgeek has quit IRC | 18:36 | |
| *** Longgeek has joined #openstack-neutron | 18:37 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron: WIP: add dependent plugin migrations https://review.openstack.org/84519 | 18:38 |
| *** sphoorti__ has quit IRC | 18:38 | |
| *** dave_tucker is now known as dave_tucker_zzz | 18:40 | |
| *** Longgeek has quit IRC | 18:42 | |
| *** jroovers has joined #openstack-neutron | 18:42 | |
| *** sbfox has joined #openstack-neutron | 18:42 | |
| *** dave_tucker_zzz is now known as dave_tucker | 18:43 | |
| mestery | marun: How can I run a specific functional test similar to how I run a specific UT test with tox? | 18:44 |
| mestery | marun: I'm writing the VXLAN version functional tests now, thus the question. :) | 18:44 |
| *** sandr8 has joined #openstack-neutron | 18:47 | |
| *** thedodd has joined #openstack-neutron | 18:48 | |
| *** Longgeek has joined #openstack-neutron | 18:53 | |
| *** tomoe_ has quit IRC | 18:55 | |
| jlibosva | mestery: I'm not maru :) , but have you tried "tox -efunctional" ? | 18:56 |
| *** tomoe_ has joined #openstack-neutron | 18:56 | |
| openstackgerrit | Aaron Rosen proposed a change to openstack/neutron: NSX: Ensure that DynamicLoop for sync doesn't raise https://review.openstack.org/84523 | 18:57 |
| *** evgenyf has quit IRC | 18:58 | |
| *** tvardeman has quit IRC | 18:58 | |
| mestery | jlibosva: I have, and that's how I run all of them, but how do I run a specific one? (And thanks for the offer of assistance!) | 18:59 |
| *** tomoe_ has quit IRC | 19:00 | |
| *** gdubreui has joined #openstack-neutron | 19:00 | |
| *** jistr has joined #openstack-neutron | 19:01 | |
| jlibosva | mestery: this worked for me: "tox -efunctional -- agent.linux.test_async_process" | 19:03 |
| ekarlso | mestery: care for a question on gre ? | 19:03 |
| HenryG | mestery: example: tox -e functional -- tests.functional.agent.linux.test_ovsdb_monitor.TestSimpleInterfaceMonitor | 19:04 |
| mestery | jilbosva: Thanks! | 19:04 |
| jlibosva | mestery: while the test is placed in "neutron/tests/functional/agent/linux/test_async_process", there is set env variable so I guess the path should be relative to "neutron/tests/functional" | 19:04 |
| ekarlso | trying to debug why dhcp traffic isn't going from the compute node to the network node | 19:05 |
| *** networkstatic has joined #openstack-neutron | 19:05 | |
| ekarlso | I can't see any dhcp packets arriving at the net node | 19:05 |
| mestery | ekarlso: Give me a few minutes, doing 3 things at once. :) | 19:05 |
| ekarlso | :) | 19:05 |
| jlibosva | happy to help :) | 19:07 |
| *** markmcclain has quit IRC | 19:08 | |
| ekarlso | dunno how to debug it | 19:08 |
| ekarlso | all seems fine :/ | 19:08 |
| *** jlibosva has quit IRC | 19:09 | |
| *** singhs has joined #openstack-neutron | 19:11 | |
| *** thuc has quit IRC | 19:12 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron: WIP: add dependent plugin migrations https://review.openstack.org/84519 | 19:12 |
| *** thuc has joined #openstack-neutron | 19:12 | |
| *** thuc has quit IRC | 19:17 | |
| *** leseb has joined #openstack-neutron | 19:21 | |
| openstackgerrit | Ian Wienand proposed a change to openstack/neutron: Record and log reason for dhcp agent resync https://review.openstack.org/81173 | 19:23 |
| openstackgerrit | Aaron Rosen proposed a change to openstack/neutron: NSX: Ensure that DynamicLoop for sync doesn't raise https://review.openstack.org/84523 | 19:29 |
| *** otherwiseguy has joined #openstack-neutron | 19:29 | |
| marun | mestery: ping | 19:33 |
| *** jp_at_hp has quit IRC | 19:33 | |
| mestery | marun: In a meeting now, FYI. | 19:34 |
| marun | mestery: I'll send an email. | 19:34 |
| ekarlso | anyone know ? | 19:34 |
| mestery | marun: Thanks! | 19:34 |
| marun | ekarlso: tcpdump is your friend :) | 19:37 |
| *** jp_at_hp has joined #openstack-neutron | 19:40 | |
| ekarlso | marun: really already done that | 19:42 |
| marun | ekarlso: so you know where the traffic is being dropped? | 19:43 |
| ekarlso | marun: well I tcpdump on the net nodes tap and nothing arrives | 19:48 |
| ekarlso | how can I tcpdump on the compute ? | 19:48 |
| ekarlso | this is IH | 19:48 |
| ekarlso | I check on the br-int bridge and I see the req there but no reponse | 19:48 |
| marun | ekarlso: you need to know each device between the source and destination and use bisection to discover at what point the traffic disappears | 19:48 |
| ekarlso | a bit crap marun to tcpdump on the phys interface of the net node as there's tons of other traffic | 19:50 |
| marun | ekarlso: filter the results for only the traffic you're looking for | 19:50 |
| *** yamahata has joined #openstack-neutron | 19:55 | |
| *** thuc has joined #openstack-neutron | 19:56 | |
| ekarlso | marun: so | 19:57 |
| ekarlso | I am tcptumping on the eth on the compute with a dst filter of the gre endpoint of the net node | 19:57 |
| ekarlso | also dumping on the endpoint on the net node and I see the dhcp packet both places | 19:58 |
| ekarlso | but when I dump on the dhcp tap interface there's no packet | 19:58 |
| *** baoli has joined #openstack-neutron | 19:58 | |
| *** baoli has quit IRC | 19:58 | |
| *** baoli has joined #openstack-neutron | 19:59 | |
| *** sbfox1 has joined #openstack-neutron | 19:59 | |
| *** zzelle has joined #openstack-neutron | 20:00 | |
| *** sbfox has quit IRC | 20:00 | |
| *** BuSerD has joined #openstack-neutron | 20:01 | |
| *** markmcclain has joined #openstack-neutron | 20:03 | |
| *** mwagner_afk has quit IRC | 20:03 | |
| *** sbfox has joined #openstack-neutron | 20:03 | |
| ekarlso | I don't have a big enough grasp on the network stuff do know where it stops next | 20:03 |
| *** sbfox1 has quit IRC | 20:03 | |
| HenryG | kevinbenton: ping | 20:05 |
| kevinbenton | HenryG: pong | 20:06 |
| marun | ekarlso: it's kind of hard to work with neutron if you don't have a networking background | 20:06 |
| ekarlso | marun: i've done it before .. | 20:06 |
| ekarlso | should the ovs-ofctl dump-flows br-int on the net node give any results except: cookie=0x0, duration=478.921s, table=0, n_packets=9, n_bytes=1434, idle_age=337, priority=0 actions=NORMAL | 20:07 |
| HenryG | kevinbenton: saw on -infra that you are looking into unit test error, Permission denied: '/var/lib/neutron'? | 20:07 |
| ekarlso | I thought it needed to have flows in order to know which gre stuff goes where | 20:08 |
| kevinbenton | HenryG: yeah | 20:08 |
| HenryG | kevinbenton: any progress? | 20:09 |
| kevinbenton | HenryG: tests are still running on my dev machine. haven't failed yet | 20:09 |
| ekarlso | marun: ? | 20:10 |
| marun | ekarlso: so you're seeing dhcp packets traverse the gre link and show up on the net node... | 20:11 |
| HenryG | kevinbenton: you trying the parallel --until-failure ? | 20:11 |
| marun | ekarlso: but then not make it to the dhcp interface? | 20:11 |
| ekarlso | marun: yes I see them ending up on the net nodes eth | 20:11 |
| kevinbenton | HenryG: yep. it just failed now | 20:11 |
| ekarlso | but shouldn't there be any flows on the compute / net node ? | 20:11 |
| ekarlso | btw | 20:12 |
| ekarlso | if I restart openvswitch on a node and then the agent | 20:12 |
| ekarlso | flows are lost... | 20:12 |
| beagles | ekarlso, you are using tunnelling? | 20:13 |
| marun | ekarlso: presuming usual config, there should be br-int and br-tun | 20:13 |
| ekarlso | beagles: GRE | 20:13 |
| beagles | ekarlso, right.. you should look at the br-tun | 20:13 |
| marun | ekarlso: is the dhcp traffic making it to br-int? | 20:13 |
| ekarlso | sometimes the ovsdb thing borks and the whole thing dies | 20:13 |
| ekarlso | when it comes back up the flows for tunneling aren't there... | 20:13 |
| beagles | !! | 20:13 |
| openstack | beagles: Error: "!" is not a valid command. | 20:13 |
| *** Sukhdev has quit IRC | 20:14 | |
| beagles | meaning that probably isn't expected behavior ;) | 20:14 |
| kevinbenton | HenryG: did you encounter the same bug? | 20:14 |
| ekarlso | sigh, I see why | 20:14 |
| ekarlso | ovsdb dies in between, one restarts it but agent doesn't pickup... | 20:15 |
| HenryG | kevinbenton: latest jenkins result on https://review.openstack.org/40296 | 20:15 |
| ekarlso | so u need to restart vswitchd and then either wait for agent to play ketchup (catchup) or restart it | 20:15 |
| kevinbenton | HenryG: ok | 20:17 |
| *** dave_tucker is now known as dave_tucker_zzz | 20:17 | |
| tchaypo | if we're talking about the thing where neutron-openvswitch-agent doesn't talk to the new ovsdb-server | 20:17 |
| tchaypo | in my experience, it doesn't catch up, not even after several days of waiting | 20:17 |
| ekarlso | https://bugs.launchpad.net/neutron/+bug/1290486 < that is a release | 20:18 |
| tchaypo | yep, that's the bug I was thinking of :) | 20:18 |
| *** sbfox has quit IRC | 20:18 | |
| ekarlso | same problem | 20:19 |
| ekarlso | I see errors about vswitchd dying | 20:19 |
| tchaypo | what i saw yesterday is that the ovsdb-client being run by n-o-a detects that the server has gone away, so it goes away | 20:20 |
| tchaypo | 30-60s later n-o-a starts a new client, but the new client doesn't seem to talk to the new server | 20:20 |
| ekarlso | root@svg-cn03:~# ps -ef | grep ovsdb-client | wc -l | 20:20 |
| ekarlso | 37 | 20:20 |
| tchaypo | I wasn't able to figure out why | 20:20 |
| ekarlso | does it keep spawning new client until vswitchd dies or smth? | 20:21 |
| ekarlso | Command: ['sudo', '/usr/bin/neutron-rootwrap', '/etc/neutron/rootwrap.conf', 'kill', '-9', '19178' - Exit code: 99 | 20:21 |
| ekarlso | uh, it seems to be missing the filter for the kill command... | 20:22 |
| *** carl_baldwin_m has joined #openstack-neutron | 20:23 | |
| tchaypo | huh. On my node I only get 4 | 20:23 |
| ekarlso | it doesn't kill the old ones.. | 20:23 |
| tchaypo | ne of those is the grep, one is the ovsdb-client process, one is neutron-rootwrap, and one is rootwrap-under-sudo | 20:23 |
| tchaypo | when i service openvswitch-switch restart (which triggers the problem), I do see those ovsdb-client processes going away | 20:24 |
| tchaypo | and being recreated about a minute later | 20:25 |
| kevinbenton | HenryG: may have found the issue. testing the fix now | 20:25 |
| *** julim has quit IRC | 20:25 | |
| ekarlso | tchaypo: yeah, and the ovs agent doesn't catch that and resync it | 20:26 |
| ekarlso | so you're kinda screwed when it happens | 20:26 |
| ekarlso | since all the flows are wiped | 20:26 |
| *** leseb has quit IRC | 20:27 | |
| *** carl_baldwin_m has quit IRC | 20:27 | |
| *** carl_baldwin_m has joined #openstack-neutron | 20:27 | |
| tchaypo | and restarting n-o-a makes it recreate the flows, right? | 20:27 |
| *** leseb has joined #openstack-neutron | 20:27 | |
| ekarlso | ye | 20:27 |
| *** Mario_ has joined #openstack-neutron | 20:28 | |
| zzelle | markmcclain, hi | 20:28 |
| ekarlso | really annoying though | 20:28 |
| jogo | got a quick neutron Q: are security group names case sensitive or not? | 20:28 |
| jogo | re: https://review.openstack.org/#/c/77347/5 | 20:28 |
| tchaypo | ekarlso: that's an understatement! | 20:29 |
| *** jorgem1 has quit IRC | 20:29 | |
| *** carl_baldwin_m has quit IRC | 20:29 | |
| *** carl_baldwin_m has joined #openstack-neutron | 20:29 | |
| ekarlso | this is freakin annoying | 20:31 |
| ekarlso | piece of scrap agent | 20:31 |
| ekarlso | traffic not hitting where it should :( | 20:31 |
| ekarlso | gawd, the ovs thing dies like a fly | 20:31 |
| *** leseb has quit IRC | 20:32 | |
| beagles | ekarlso: out of curiousity what OS and ovs version are you using? | 20:33 |
| ekarlso | ubuntu 14.04, ovs 2.0 or so | 20:33 |
| markmcclain | zzelle: hi | 20:34 |
| zzelle | markmclain, do you have some minutes to speek about https://blueprints.launchpad.net/neutron/+spec/provider-network-partial-specs ? | 20:35 |
| *** jecarey has quit IRC | 20:35 | |
| tchaypo | I'm also ubuntu 14.04, ovsdb-server reports "1.10.2 Compiled Sep 23 2013" | 20:35 |
| markmcclain | zzelle: currently in the TC meeting | 20:35 |
| zzelle | markmcclain, i will wait for the meeting end | 20:36 |
| *** carl_baldwin_m has quit IRC | 20:36 | |
| *** carl_baldwin_m has joined #openstack-neutron | 20:37 | |
| *** jistr has quit IRC | 20:37 | |
| beagles | ah, life on the bleeding edge.. unfortunately, there is often a lot of bleeding there ;) | 20:37 |
| carl_baldwin | marun: Is there a bug already filed for the problem we discussed with rpc workers? | 20:39 |
| mestery | ubuntu 14.04 is only using OVS 1.10? | 20:41 |
| ekarlso | no | 20:41 |
| ekarlso | it has ovs 2.0 | 20:41 |
| * mestery breathes a sigh of relief. | 20:41 | |
| tchaypo | I wonder why I have such an old version. | 20:43 |
| *** jroovers has quit IRC | 20:43 | |
| *** carl_baldwin_m has quit IRC | 20:44 | |
| *** dims_ has quit IRC | 20:44 | |
| *** carl_baldwin_m has joined #openstack-neutron | 20:44 | |
| ekarlso | ok | 20:45 |
| ekarlso | I just rebooted everything and now it works... | 20:45 |
| *** Longgeek has quit IRC | 20:45 | |
| ekarlso | evil neutrons | 20:45 |
| *** sbfox has joined #openstack-neutron | 20:45 | |
| *** jecarey has joined #openstack-neutron | 20:46 | |
| ekarlso | now let's see how long it takes before it dies agian... | 20:48 |
| marun | carl_baldwin: ah, no. | 20:49 |
| carl_baldwin | That is okay. I can file one. | 20:49 |
| *** sbfox has quit IRC | 20:49 | |
| *** sballe has joined #openstack-neutron | 20:49 | |
| marun | carl_baldwin: danke | 20:49 |
| *** sbfox has joined #openstack-neutron | 20:49 | |
| beagles | ekarlso, it would be cool if you could capture the info around ovs failing and pass it on to the openvswitch crowd. | 20:49 |
| beagles | ekarlso, while neutron could handle recovery better, ovs crashing is a pretty serious situation and should also be resolved. | 20:51 |
| *** carl_baldwin_m has quit IRC | 20:51 | |
| *** carl_baldwin_m has joined #openstack-neutron | 20:51 | |
| ekarlso | beagles: where do I check ? | 20:52 |
| beagles | ekarlso, the open vswitch logs might be a good place to start /var/log/openvswitch | 20:53 |
| beagles | ekarlso, anything that might be related in messages | 20:53 |
| *** crc32 has joined #openstack-neutron | 20:53 | |
| *** leseb has joined #openstack-neutron | 20:54 | |
| tchaypo | ekarlso: can you update the bug with anything you find/provide? I'll try to grab matching info as well | 20:54 |
| ekarlso | tchaypo: sure | 20:58 |
| tchaypo | I'll also try to poke my system into having a less ancient version of OVS and see if I can still replicate | 20:58 |
| *** carl_baldwin_m has quit IRC | 20:58 | |
| jogo | markmcclain: when you have a minute https://review.openstack.org/#/c/77347/5 | 20:58 |
| *** carl_baldwin_m has joined #openstack-neutron | 20:59 | |
| *** blogan has joined #openstack-neutron | 20:59 | |
| marun | carl_baldwin: can you please send me a bug link when you have one? | 21:00 |
| carl_baldwin | Yes, I will. | 21:00 |
| markmcclain | jogo: +1 | 21:00 |
| *** blogan has quit IRC | 21:01 | |
| jogo | markmcclain: so case-insensitive neutron security groups | 21:01 |
| jogo | this is another case of nova-network neutron tiny diffs AHH | 21:01 |
| jogo | markmcclain: and I don't want to make any changes without double checking and triplechecking | 21:02 |
| markmcclain | yeah understood | 21:02 |
| jogo | markmcclain: so two concerns -- this is a slightly incompatible behavioral change. and 2 what does neutron *want* to do | 21:02 |
| markmcclain | my pref is case-sensitive long term, but we've to preserve behavior | 21:03 |
| markmcclain | with nova | 21:04 |
| *** manishg has quit IRC | 21:04 | |
| jogo | so nova does both apparently | 21:04 |
| *** zhipeng has joined #openstack-neutron | 21:05 | |
| jogo | markmcclain: I guess I see no good answer here yet | 21:05 |
| *** carl_baldwin_m has quit IRC | 21:06 | |
| jogo | you are case sensitive, your API is stable. we have nova-net case insensitive and our API is stable | 21:06 |
| *** carl_baldwin_m has joined #openstack-neutron | 21:06 | |
| *** singhs has quit IRC | 21:06 | |
| jogo | russellb: ^ | 21:07 |
| markmcclain | yeah not sure what the correct behavior should be other to than preserve the behavior based on endpoint used | 21:07 |
| *** casmith32 has quit IRC | 21:08 | |
| russellb | add an API extension that allows you to specify "case_insensistive" via a query param or something | 21:08 |
| * russellb thinking off the cuff | 21:08 | |
| jogo | russellb: either way you break an APIs default behavior by doing that | 21:08 |
| jogo | either nova or neutron | 21:08 |
| jogo | I think we have to bite the bullet somewhere. I do like the idea of adding a case_insensitive option for the API we are forced to change | 21:09 |
| jogo | so its a fairly minor break | 21:09 |
| jogo | although I think the case sensitive would be per tenant | 21:10 |
| jogo | not per call | 21:10 |
| *** sbfox has quit IRC | 21:13 | |
| *** carl_baldwin_m has quit IRC | 21:13 | |
| *** carl_baldwin_m has joined #openstack-neutron | 21:13 | |
| jogo | markmcclain: anyway -1ed the nova patch asking josh to bring this to the ML as this is just ugly and painful | 21:13 |
| jogo | markmcclain: do you know of any other issues like this? | 21:13 |
| jogo | markmcclain: maybe neutron needs a nova compat mode (set per tenant?) | 21:14 |
| jogo | assuming other minor differences like this will arise | 21:14 |
| salv-orlando | make sense. Honestly I'm not conviced by the case_sensitive query params. Maybe neutron could just move and became case-insensitive on sec group names? | 21:15 |
| markmcclain | zzelle: ping | 21:15 |
| markmcclain | salv-orlando: might be way to address it | 21:15 |
| *** sridhar has quit IRC | 21:15 | |
| salv-orlando | on another note, have we already filed a bug for neutron UT failure re: access to /var/lib/neutron? | 21:15 |
| zzelle | markmcclain, pong | 21:15 |
| jogo | salv-orlando: IMHO that is a slightly backwards incompatible change. but one we may have to live with | 21:15 |
| markmcclain | honestly seems kind of crazy someone would define a "Default" and "default" group | 21:16 |
| jogo | markmcclain: heh | 21:16 |
| salv-orlando | markmcclain: I am convinced it will address that. I am not convinced it is the right way of doing that. | 21:16 |
| jogo | and dEfAuLt | 21:16 |
| *** mlavalle has joined #openstack-neutron | 21:16 | |
| zzelle | zzelle, my question(s) what are the next steps for https://blueprints.launchpad.net/neutron/+spec/provider-network-partial-specs ? to get it approved | 21:17 |
| zzelle | markmclain, it has been approved during ml2 meeting | 21:17 |
| jogo | markmcclain: do you know of any other issues like this off hand? | 21:17 |
| markmcclain | zzelle: I'll be sending out an email later formally opening for Juno | 21:17 |
| markmcclain | that change will include requirement to file blueprints in the new format | 21:18 |
| *** rkukura has quit IRC | 21:18 | |
| zzelle | markmcclain, new format ? like using gerrit for BP as for nova ? | 21:18 |
| markmcclain | that is the next step, but I've been trying to keep the churn in the tree down for a day or so should the RC have any issues | 21:18 |
| markmcclain | zzelle: exactly | 21:18 |
| salv-orlando | markmcclain, jogo: this has been happening for the past 4 hours -> http://bit.ly/1lAFeUb | 21:18 |
| salv-orlando | hitting gate as well | 21:18 |
| salv-orlando | are you already aware? | 21:19 |
| zzelle | markmcclain, interested to see how it evolves | 21:19 |
| *** jp_at_hp has quit IRC | 21:19 | |
| jogo | salv-orlando: ahh that is just on neutron unit tests | 21:19 |
| markmcclain | zzelle: yeah, but it has to be better than the old way where we had blueprints in 6 different mediums structured lots of ways | 21:19 |
| jogo | salv-orlando: my guess is out of memory | 21:19 |
| jogo | salv-orlando: lets hope over to infra | 21:20 |
| jogo | salv-orlando: do you have an e-r patch up for this? | 21:20 |
| markmcclain | salv-orlando: this merged: https://review.openstack.org/#/c/83474/ | 21:20 |
| *** carl_baldwin_m has quit IRC | 21:20 | |
| *** jprovazn has quit IRC | 21:20 | |
| salv-orlando | I can have it in 5 minutes | 21:20 |
| *** carl_baldwin_m has joined #openstack-neutron | 21:20 | |
| jogo | salv-orlando: great | 21:21 |
| salv-orlando | markmcclain: what has 83474 to do with the unit test failure? | 21:21 |
| markmcclain | oh wait nevermind its not related to keystone change | 21:21 |
| ekarlso | is it normal to get alot of tcp checksum incorrect on instances ? | 21:21 |
| *** foexle has quit IRC | 21:22 | |
| *** carl_baldwin_m has quit IRC | 21:27 | |
| *** carl_baldwin_m has joined #openstack-neutron | 21:28 | |
| *** catohornet has quit IRC | 21:30 | |
| *** bjornar has quit IRC | 21:30 | |
| *** sbfox has joined #openstack-neutron | 21:33 | |
| *** carl_baldwin_m has quit IRC | 21:35 | |
| *** carl_baldwin_m has joined #openstack-neutron | 21:35 | |
| *** thuc has quit IRC | 21:41 | |
| *** thuc has joined #openstack-neutron | 21:41 | |
| *** carl_baldwin_m has quit IRC | 21:42 | |
| *** carl_baldwin_m has joined #openstack-neutron | 21:42 | |
| *** sbfox has quit IRC | 21:43 | |
| *** sbfox has joined #openstack-neutron | 21:44 | |
| carl_baldwin | marun: https://bugs.launchpad.net/neutron/+bug/1301035 | 21:46 |
| *** BuSerD has quit IRC | 21:46 | |
| *** thuc has quit IRC | 21:46 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron: Pin to lower oslo.config version https://review.openstack.org/84562 | 21:46 |
| *** carl_baldwin_m has quit IRC | 21:49 | |
| *** salv-orlando_ has joined #openstack-neutron | 21:49 | |
| *** enikanorov has joined #openstack-neutron | 21:49 | |
| *** BuSerD has joined #openstack-neutron | 21:49 | |
| *** carl_baldwin_m has joined #openstack-neutron | 21:49 | |
| *** francois_eleouet has joined #openstack-neutron | 21:52 | |
| *** dguerri_ has joined #openstack-neutron | 21:53 | |
| *** cgoncalv1s has joined #openstack-neutron | 21:54 | |
| *** dguerri has quit IRC | 21:54 | |
| *** openstackgerrit has quit IRC | 21:54 | |
| *** salv-orlando has quit IRC | 21:54 | |
| *** markvoelker has quit IRC | 21:54 | |
| *** feleouet has quit IRC | 21:54 | |
| *** enikanorov_ has quit IRC | 21:54 | |
| *** cgoncalves has quit IRC | 21:54 | |
| *** salv-orlando_ is now known as salv-orlando | 21:54 | |
| *** singhs has joined #openstack-neutron | 21:56 | |
| *** carl_baldwin_m has quit IRC | 21:56 | |
| *** carl_baldwin_m has joined #openstack-neutron | 21:57 | |
| markmcclain | kevinbenton: we can't pin the version | 21:57 |
| markmcclain | so we need to find where the conflict is occurring | 21:57 |
| *** dvorkinista has quit IRC | 21:57 | |
| *** carl_baldwin_m has quit IRC | 21:57 | |
| kevinbenton | markmcclain: the problem is now it's being evaluated immediately on definition now | 21:58 |
| kevinbenton | markmcclain: will have to change several unit tests. is that okay? | 21:58 |
| markmcclain | yes that is preferrable to fix the tests | 21:59 |
| markmcclain | because we'll have to fix long term anyway | 21:59 |
| kevinbenton | markmcclain: ok. i will do it then | 22:00 |
| markmcclain | thanks for working on this | 22:00 |
| markmcclain | I've got to step out for a bit, but I'll check in when I return | 22:00 |
| *** rkukura has joined #openstack-neutron | 22:01 | |
| markmcclain | fortunately this is only breaking UT, so we won't be blocking the gate | 22:01 |
| jogo | markmcclain: I am working fixing the test | 22:01 |
| jogo | markmcclain: will have a patch soon | 22:02 |
| markmcclain | jogo: oh cool | 22:02 |
| markmcclain | kevinbenton: ^^ | 22:02 |
| jogo | markmcclain: may take a minute because forgot to drop the '-r' in tox | 22:02 |
| kevinbenton | jogo: it's more than one | 22:03 |
| kevinbenton | jogo: like 7 of them | 22:03 |
| *** sungju_ has joined #openstack-neutron | 22:04 | |
| jogo | kevinbenton: I run devpi-server | 22:04 |
| jogo | so its just a minute | 22:04 |
| jogo | https://pypi.python.org/pypi/devpi-server/1.2.1 | 22:04 |
| jogo | its a great trick | 22:06 |
| *** dims_ has joined #openstack-neutron | 22:06 | |
| *** armax has quit IRC | 22:07 | |
| jogo | kevinbenton markmcclain: fix https://review.openstack.org/84568 | 22:10 |
| markmcclain | jogo: cool | 22:11 |
| *** banix has quit IRC | 22:13 | |
| markmcclain | jogo: looks like we lost the lottery and the py26 check will take ~hr because of the node it ended up on | 22:14 |
| *** leseb has quit IRC | 22:14 | |
| *** WackoRobie has quit IRC | 22:14 | |
| jogo | markmcclain: trying to propose a match to milestone proposed but hitting something funny | 22:14 |
| *** alexpilotti has quit IRC | 22:14 | |
| *** leseb has joined #openstack-neutron | 22:15 | |
| *** sballe has quit IRC | 22:15 | |
| *** WackoRobie has joined #openstack-neutron | 22:15 | |
| jogo | ahh there we go defaultbranch=stable/havana | 22:18 |
| *** dims_ has quit IRC | 22:18 | |
| jogo | err https://review.openstack.org/84569 | 22:18 |
| *** leseb has quit IRC | 22:19 | |
| *** WackoRobie has quit IRC | 22:19 | |
| *** jobewan has quit IRC | 22:21 | |
| *** jgrimm has quit IRC | 22:23 | |
| *** jaypipes has joined #openstack-neutron | 22:24 | |
| *** djoreilly has quit IRC | 22:24 | |
| *** zhipeng has quit IRC | 22:24 | |
| *** zhipeng has joined #openstack-neutron | 22:24 | |
| *** thuc has joined #openstack-neutron | 22:25 | |
| *** thuc has quit IRC | 22:25 | |
| *** thuc has joined #openstack-neutron | 22:26 | |
| *** changbl has quit IRC | 22:32 | |
| markmcclain | jogo: thanks I'll let ttx know that we'll need an RC2 | 22:33 |
| jogo | markmcclain: cool | 22:34 |
| *** dims_ has joined #openstack-neutron | 22:36 | |
| jogo | I'm happy the fix was so easy | 22:38 |
| jogo | but this brings up an interesting point | 22:38 |
| jogo | sdague: ^ we should use integrated gate reqs for unit tests too ? | 22:38 |
| *** prad has quit IRC | 22:38 | |
| arosen | jogo: awesome thanks for that fix. I've been changing /var/run/neutron to /opt/stack/neutron for a while now | 22:40 |
| * arosen manually in the code X.o | 22:40 | |
| *** thuc has quit IRC | 22:42 | |
| jogo | arosen: haha | 22:42 |
| *** thuc has joined #openstack-neutron | 22:42 | |
| arosen | jogo: https://review.openstack.org/#/c/75253/2/neutron/common/config.py :) | 22:43 |
| *** thedodd has quit IRC | 22:43 | |
| *** yfried__ has quit IRC | 22:44 | |
| jogo | arosen: hehe | 22:44 |
| *** otherwiseguy has quit IRC | 22:44 | |
| jogo | arosen markmcclain so maybe a better solution is to use a tempdir for state_path itself | 22:45 |
| jogo | markmcclain: thoughts? | 22:45 |
| *** baoli has quit IRC | 22:46 | |
| *** manishg has joined #openstack-neutron | 22:46 | |
| *** thuc has quit IRC | 22:46 | |
| markmcclain | yes state_path should probably be a tempdir | 22:47 |
| *** zzelle has quit IRC | 22:48 | |
| *** sn6i23a has quit IRC | 22:51 | |
| *** markmcclain has quit IRC | 22:52 | |
| jogo | markmcclain: matches revised | 22:52 |
| tchaypo | ekarlso: I expect you saw a massive email from me :) | 22:52 |
| *** mlavalle has quit IRC | 22:53 | |
| ekarlso | which one tchaypo ? | 22:54 |
| ekarlso | I get tons : | 22:54 |
| *** jorgem has joined #openstack-neutron | 22:56 | |
| *** dvorkinista has joined #openstack-neutron | 22:56 | |
| *** thuc has joined #openstack-neutron | 22:57 | |
| *** armax has joined #openstack-neutron | 22:59 | |
| *** thuc has joined #openstack-neutron | 22:59 | |
| *** armax has quit IRC | 22:59 | |
| *** dvorkinista has quit IRC | 23:01 | |
| *** jorgem has quit IRC | 23:03 | |
| *** jorgem has joined #openstack-neutron | 23:04 | |
| *** nplanel_ has joined #openstack-neutron | 23:05 | |
| tchaypo | haha | 23:07 |
| tchaypo | the one about n-o-a not replacing flows when ovsdb-server restarts | 23:07 |
| *** yamahata has quit IRC | 23:08 | |
| *** crc32 has quit IRC | 23:08 | |
| *** jorgem has quit IRC | 23:08 | |
| *** WackoRobie has joined #openstack-neutron | 23:08 | |
| lifeless | tchaypo: was that to -dev? | 23:13 |
| *** pcm_ has quit IRC | 23:14 | |
| *** networkstatic is now known as networkstatic_zZ | 23:14 | |
| *** jorgem has joined #openstack-neutron | 23:16 | |
| tchaypo | no, I updated the bug | 23:18 |
| *** jorgem has quit IRC | 23:21 | |
| *** _afazekas has quit IRC | 23:21 | |
| *** dvorkinista has joined #openstack-neutron | 23:23 | |
| *** thuc has quit IRC | 23:30 | |
| *** mwagner_afk has joined #openstack-neutron | 23:32 | |
| *** pcm_ has joined #openstack-neutron | 23:33 | |
| *** otherwiseguy has joined #openstack-neutron | 23:36 | |
| marun | carl_baldwin: danke (for the bug link) | 23:40 |
| carl_baldwin | pw | 23:40 |
| carl_baldwin | yw | 23:40 |
| *** kevinbenton has left #openstack-neutron | 23:44 | |
| *** thuc has joined #openstack-neutron | 23:45 | |
| *** kevinbenton has joined #openstack-neutron | 23:45 | |
| marun | carl_baldwin: is the patch ready for review or are you still iterating? | 23:46 |
| carl_baldwin | I think patch set 2 is ready for review. | 23:47 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!