| *** abaindur_ has joined #openstack-lbaas | 00:07 | |
| *** abaindur has quit IRC | 00:07 | |
| *** abaindur has joined #openstack-lbaas | 00:10 | |
| *** abaindur_ has quit IRC | 00:12 | |
| *** abaindur has quit IRC | 00:17 | |
| *** abaindur has joined #openstack-lbaas | 00:17 | |
| *** abaindur has quit IRC | 00:20 | |
| *** abaindur has joined #openstack-lbaas | 00:20 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 00:25 |
|---|---|---|
| *** ivve has quit IRC | 00:27 | |
| *** luksky has quit IRC | 00:36 | |
| *** yamamoto has joined #openstack-lbaas | 00:46 | |
| *** yamamoto has quit IRC | 00:50 | |
| *** abaindur_ has joined #openstack-lbaas | 00:59 | |
| *** abaindur has quit IRC | 01:01 | |
| *** ricolin has joined #openstack-lbaas | 01:07 | |
| *** bzhao__ has joined #openstack-lbaas | 01:11 | |
| *** yamamoto has joined #openstack-lbaas | 01:17 | |
| *** yamamoto has quit IRC | 01:23 | |
| *** abaindur_ has quit IRC | 01:53 | |
| *** hongbin has joined #openstack-lbaas | 02:34 | |
| *** yamamoto has joined #openstack-lbaas | 02:37 | |
| *** yamamoto has quit IRC | 02:42 | |
| *** ramishra has joined #openstack-lbaas | 03:11 | |
| *** yamamoto has joined #openstack-lbaas | 03:23 | |
| *** yamamoto has quit IRC | 03:27 | |
| *** yamamoto has joined #openstack-lbaas | 03:30 | |
| *** yamamoto has quit IRC | 04:10 | |
| *** yamamoto has joined #openstack-lbaas | 04:10 | |
| *** hongbin has quit IRC | 04:10 | |
| *** abaindur has joined #openstack-lbaas | 04:15 | |
| *** abaindur has quit IRC | 04:22 | |
| *** abaindur has joined #openstack-lbaas | 04:28 | |
| *** abaindur has quit IRC | 04:29 | |
| *** ricolin is now known as ricolin_phone | 05:02 | |
| *** ricolin_phone is now known as ricolin | 05:04 | |
| *** ricolin is now known as ricolin_ | 05:04 | |
| *** ricolin_ is now known as ricolin_afk | 05:04 | |
| *** ricolin_afk is now known as ricolin | 05:05 | |
| *** yamamoto has quit IRC | 05:09 | |
| openstackgerrit | Reedip proposed openstack/octavia-tempest-plugin master: Support skipping SESSION_PERSISTENCE if L7 Protocol is disabled https://review.openstack.org/645495 | 05:41 |
| *** yboaron has joined #openstack-lbaas | 05:44 | |
| *** yboaron_ has joined #openstack-lbaas | 05:58 | |
| *** yboaron has quit IRC | 06:01 | |
| *** AlexStaf has quit IRC | 06:31 | |
| *** abaindur has joined #openstack-lbaas | 06:50 | |
| *** ivve has joined #openstack-lbaas | 06:58 | |
| *** ccamposr has joined #openstack-lbaas | 07:09 | |
| *** pcaruana has joined #openstack-lbaas | 07:10 | |
| *** abaindur has quit IRC | 07:18 | |
| *** ramishra_ has joined #openstack-lbaas | 07:31 | |
| *** gthiemonge has joined #openstack-lbaas | 07:33 | |
| *** ramishra has quit IRC | 07:34 | |
| *** abaindur has joined #openstack-lbaas | 07:37 | |
| *** gcheresh has joined #openstack-lbaas | 07:38 | |
| *** luksky has joined #openstack-lbaas | 07:57 | |
| *** rpittau|afk is now known as rpittau | 08:06 | |
| *** celebdor has joined #openstack-lbaas | 08:13 | |
| *** AlexStaf has joined #openstack-lbaas | 08:19 | |
| *** abaindur has quit IRC | 08:23 | |
| *** rcernin has quit IRC | 08:36 | |
| openstackgerrit | Nir Magnezi proposed openstack/neutron-lbaas stable/queens: Choose correct log option by listener protocol https://review.openstack.org/647689 | 09:09 |
| openstackgerrit | Nir Magnezi proposed openstack/neutron-lbaas stable/queens: Modify a detail in parse_stat https://review.openstack.org/647690 | 09:12 |
| *** yboaron_ has quit IRC | 09:23 | |
| *** luksky has quit IRC | 09:35 | |
| *** velizarx has joined #openstack-lbaas | 09:39 | |
| *** rcernin has joined #openstack-lbaas | 09:44 | |
| *** yboaron_ has joined #openstack-lbaas | 10:10 | |
| *** yboaron_ has quit IRC | 10:16 | |
| *** yboaron_ has joined #openstack-lbaas | 10:16 | |
| *** rcernin has quit IRC | 10:24 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia stable/queens: Fix a systemd race condition in the amphora https://review.openstack.org/647708 | 10:34 |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 10:38 |
| *** psachin has joined #openstack-lbaas | 10:51 | |
| *** luksky has joined #openstack-lbaas | 11:02 | |
| *** ricolin has quit IRC | 11:04 | |
| *** trown|outtypewww is now known as trown | 12:08 | |
| *** ramishra_ has quit IRC | 13:59 | |
| *** ramishra has joined #openstack-lbaas | 14:15 | |
| openstackgerrit | Nir Magnezi proposed openstack/octavia master: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647752 | 14:27 |
| *** yboaron_ has quit IRC | 15:26 | |
| *** trown is now known as trown|lunch | 15:47 | |
| *** ricolin has joined #openstack-lbaas | 16:04 | |
| *** gcheresh has quit IRC | 16:08 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 16:13 |
| *** celebdor has quit IRC | 16:13 | |
| *** ccamposr has quit IRC | 16:18 | |
| *** psachin has quit IRC | 16:20 | |
| cgoncalves | johnsom, missing ubuntu@{$AMP_IP} | 16:24 |
| johnsom | Doh | 16:24 |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 16:25 |
| johnsom | Just trying to prove out my theory since we are still getting the same result | 16:25 |
| cgoncalves | should amphora-agent.service has a "Requires: amphora-netns" ? | 16:25 |
| cgoncalves | Requires={{ amphora_netns }}.service | 16:26 |
| johnsom | Well, this is the oddness of how amphora-netns service was setup. I think amphora-netns is written out by the amphora agent | 16:27 |
| cgoncalves | true | 16:28 |
| *** AlexStaf has quit IRC | 16:28 | |
| johnsom | I'm not sure it really needs to be that way. I think it could be moved to the element since it's not really dependent on any of the amp agent stuff. I mean we can make the namespace name a "known value" | 16:28 |
| johnsom | Let | 16:29 |
| johnsom | Let's see if my theory is right with this patch run, then decided what needs to happen here. | 16:29 |
| johnsom | My guess is the plugged_interfaces file is getting written out just before the amphora-netns starts, but before the agent actually starts trying to bring them up | 16:30 |
| *** fnaval has joined #openstack-lbaas | 16:37 | |
| *** ramishra has quit IRC | 16:37 | |
| cgoncalves | why now? why only on rocky grenade job? why can't we reproduce? so many unanswered questions :( | 16:41 |
| johnsom | Yes, agreed. I have seen random gate issues that are similar to this, just not nearly as reproducible. | 16:43 |
| *** mloza has joined #openstack-lbaas | 16:44 | |
| cgoncalves | amphora-netns reports "Cannot create namespace file "/var/run/netns/amphora-haproxy": File exists" at boot time | 16:46 |
| cgoncalves | anything else creating the namespace? | 16:46 |
| *** jiteka1 has joined #openstack-lbaas | 16:55 | |
| *** luksky has quit IRC | 16:57 | |
| johnsom | Yeah, I think that is related. The amphora-netns does as well as the amphora-agent | 16:58 |
| johnsom | A reboot can also trigger that | 16:58 |
| *** trown|lunch is now known as trown | 16:58 | |
| johnsom | Oh, not that message, but a related /etc directory is present | 16:58 |
| *** rpittau is now known as rpittau|afk | 16:59 | |
| *** ricolin has quit IRC | 17:13 | |
| cgoncalves | eth2 not in amphora-netns | 17:38 |
| *** ivve has quit IRC | 17:39 | |
| johnsom | Blah, so it's another RTNETLINK confusing error.... | 17:39 |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 17:52 |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 17:53 |
| *** yboaron_ has joined #openstack-lbaas | 19:07 | |
| *** velizarx has quit IRC | 19:11 | |
| *** velizarx has joined #openstack-lbaas | 19:17 | |
| *** abaindur has joined #openstack-lbaas | 19:19 | |
| *** luksky has joined #openstack-lbaas | 19:21 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 19:25 |
| *** velizarx has quit IRC | 19:33 | |
| openstackgerrit | Nir Magnezi proposed openstack/octavia stable/rocky: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647895 | 19:36 |
| openstackgerrit | Nir Magnezi proposed openstack/octavia stable/rocky: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647895 | 19:36 |
| openstackgerrit | Nir Magnezi proposed openstack/octavia stable/rocky: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647895 | 19:38 |
| openstackgerrit | Nir Magnezi proposed openstack/octavia stable/queens: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647896 | 19:39 |
| openstackgerrit | Nir Magnezi proposed openstack/octavia stable/queens: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647896 | 19:41 |
| openstackgerrit | Nir Magnezi proposed openstack/octavia stable/queens: Adds server_certs_key_passphrase to octavia.conf https://review.openstack.org/647896 | 19:41 |
| nmagnezi | </noise> | 19:41 |
| *** velizarx has joined #openstack-lbaas | 20:07 | |
| *** velizarx has quit IRC | 20:13 | |
| *** abaindur has quit IRC | 20:55 | |
| *** abaindur has joined #openstack-lbaas | 21:05 | |
| *** trown is now known as trown|outtypewww | 21:07 | |
| *** pcaruana has quit IRC | 21:33 | |
| *** yboaron_ has quit IRC | 21:35 | |
| cgoncalves | where the hell is ens7/eth2... not in the root namespace nor in the amphora-haproxy namespace | 21:46 |
| johnsom | Right, my question as well. I'm doing some local tests to see where I *should* see it. | 21:48 |
| johnsom | We see confirmation that the kernel renamed it to eth2 in the syslog, but after that it's missing | 21:49 |
| cgoncalves | yep | 21:51 |
| johnsom | Yeah, so local it is immediately in the ip link list | 21:52 |
| cgoncalves | https://github.com/openstack/octavia/blob/d138f3c0f784f7b302ef8bbf84d3318e8b33fbe8/octavia/amphorae/backends/agent/api_server/plug.py#L214-L219 | 21:52 |
| johnsom | https://www.irccloud.com/pastebin/HYEg7FMq/ | 21:52 |
| cgoncalves | ^ is what moves the interface to the namespace. it doesn't error there so... :/ | 21:52 |
| johnsom | Yeah, plus we have been doing that forever. I also looked, it's pulling the latest version of pyroute2 in as well. | 21:53 |
| johnsom | The other oddness is, in the code we ifdown then ifup, but the ifdown doesn't seem to error either | 21:53 |
| * cgoncalves nods | 21:54 | |
| johnsom | Yeah, this is *neat* | 21:56 |
| *** abaindur has quit IRC | 22:03 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/queens: Fix a systemd race condition in the amphora https://review.openstack.org/647708 | 22:15 |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 22:16 |
| johnsom | Ok, I am going to run an errand, will be back in an hour or so. | 22:16 |
| cgoncalves | same here, back in 8 hours or so :) | 22:17 |
| cgoncalves | not that it would help much, but one idea if those error log messages you just added turn out not being useful would be to switch the amphora to bionic | 22:21 |
| *** abaindur has joined #openstack-lbaas | 22:36 | |
| *** rcernin has joined #openstack-lbaas | 22:38 | |
| *** abaindur has quit IRC | 23:05 | |
| *** abaindur has joined #openstack-lbaas | 23:06 | |
| lxkong | johnsom: hi, is there anything needs to be done for this patch https://review.openstack.org/#/c/647628? | 23:06 |
| johnsom | lxkong: I was waiting for your ok. Then I will do a release note. | 23:09 |
| *** fnaval has quit IRC | 23:13 | |
| *** fnaval has joined #openstack-lbaas | 23:19 | |
| colin- | "insert_headers" is onlyl respected by listeners with protocol HTTP/HTTPS right? | 23:26 |
| lxkong | johnsom: i replied yersterday :-) | 23:26 |
| johnsom | colin-: Correct | 23:27 |
| lxkong | colin-: and also pool with HTTP(S) i think | 23:27 |
| colin- | thanks | 23:28 |
| *** luksky has quit IRC | 23:28 | |
| johnsom | The only TCP think like it would be using the proxy protocol, but that is configured differently | 23:30 |
| colin- | understood. while i am bothering you folks, which listener protocols are supported for pools of proxy protocol? all? | 23:31 |
| johnsom | Proxy protocol can only be set on pools. It is it's own protocol which is basically a TCP protocol that injects a prefix to the TCP stream | 23:32 |
| colin- | understood, but want to be really specific with which listener types are valid when pool type is proxy | 23:32 |
| johnsom | Ah, TCP, HTTPS (pass through), ... Otherwise I'm not sure | 23:33 |
| johnsom | There is a patch that builds that table, but it isn't approved yet | 23:34 |
| johnsom | colin- http://logs.openstack.org/40/594040/38/check/build-openstack-api-ref/aa9bfe6/html/v2/index.html?expanded=create-pool-detail#valid-protocol-combinations | 23:35 |
| johnsom | However, I don't know that the grid has been validated yet. It's not an approved patch | 23:36 |
| colin- | damn right there in the doc i've been scrolling past | 23:37 |
| colin- | thanks that helps a lot | 23:37 |
| openstackgerrit | Michael Johnson proposed openstack/python-octaviaclient master: Make sure we always requests JSON responses https://review.openstack.org/647628 | 23:43 |
| johnsom | lxkong https://review.openstack.org/647628 Added the release note, so you can re-add your +1. I will look into a release, just not sure if I can given it is the end of the Stein cycle. | 23:44 |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/queens: Fix a systemd race condition in the amphora https://review.openstack.org/647708 | 23:53 |
| openstackgerrit | Michael Johnson proposed openstack/octavia stable/rocky: DNM: Gate check https://review.openstack.org/628007 | 23:54 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!