| *** iyamahat__ has joined #openstack-neutron | 00:01 | |
| *** njohnston_ has joined #openstack-neutron | 00:02 | |
| *** njohnston has quit IRC | 00:03 | |
| *** iyamahat_ has quit IRC | 00:04 | |
| *** cshen has joined #openstack-neutron | 00:27 | |
| *** cshen has quit IRC | 00:31 | |
| *** zhanglong has quit IRC | 00:32 | |
| *** cmart has quit IRC | 00:33 | |
| *** iyamahat__ has quit IRC | 00:41 | |
| *** iyamahat__ has joined #openstack-neutron | 00:41 | |
| *** cheng1 has joined #openstack-neutron | 00:54 | |
| *** ileixe has joined #openstack-neutron | 00:56 | |
| *** dsneddon has quit IRC | 01:07 | |
| *** igordc has quit IRC | 01:16 | |
| openstackgerrit | Frank Wang proposed openstack/ovsdbapp master: add an option to let the user choose the right time to start connection https://review.openstack.org/644502 | 01:18 |
|---|---|---|
| *** njohnston_ has quit IRC | 01:19 | |
| *** imacdonn has quit IRC | 01:19 | |
| *** mriedem has quit IRC | 01:26 | |
| *** dsneddon has joined #openstack-neutron | 01:35 | |
| *** dsneddon has quit IRC | 01:39 | |
| *** tbachman has joined #openstack-neutron | 01:40 | |
| *** mattw4 has quit IRC | 01:42 | |
| *** dsneddon has joined #openstack-neutron | 02:09 | |
| *** dsneddon has quit IRC | 02:14 | |
| *** yamahata has quit IRC | 02:18 | |
| *** iyamahat__ has quit IRC | 02:18 | |
| *** cshen has joined #openstack-neutron | 02:27 | |
| *** cshen has quit IRC | 02:32 | |
| *** dsneddon has joined #openstack-neutron | 02:34 | |
| *** dsneddon has quit IRC | 02:40 | |
| *** markvoelker has joined #openstack-neutron | 02:43 | |
| *** janki has quit IRC | 02:44 | |
| *** binh_ has quit IRC | 02:51 | |
| *** binh_ has joined #openstack-neutron | 02:51 | |
| *** dsneddon has joined #openstack-neutron | 02:52 | |
| *** wolverineav has joined #openstack-neutron | 02:55 | |
| *** dsneddon has quit IRC | 02:59 | |
| *** yamamoto has joined #openstack-neutron | 03:01 | |
| *** cheng1 has quit IRC | 03:19 | |
| *** dsneddon has joined #openstack-neutron | 03:26 | |
| *** iyamahat has joined #openstack-neutron | 03:27 | |
| *** dsneddon has quit IRC | 03:41 | |
| *** nicolasbock has quit IRC | 03:53 | |
| *** cheng1 has joined #openstack-neutron | 03:56 | |
| *** janki has joined #openstack-neutron | 04:00 | |
| *** yamamoto has quit IRC | 04:03 | |
| *** dsneddon has joined #openstack-neutron | 04:07 | |
| *** whoami-rajat has joined #openstack-neutron | 04:10 | |
| *** dsneddon has quit IRC | 04:12 | |
| openstackgerrit | Frank Wang proposed openstack/ovsdbapp master: add an option to let the user choose the right time to start connection https://review.openstack.org/644502 | 04:13 |
| *** betherly has joined #openstack-neutron | 04:19 | |
| *** ileixe has quit IRC | 04:21 | |
| *** betherly has quit IRC | 04:24 | |
| *** ileixe has joined #openstack-neutron | 04:34 | |
| *** dsneddon has joined #openstack-neutron | 04:35 | |
| *** dsneddon has quit IRC | 04:40 | |
| *** lpetrut has joined #openstack-neutron | 04:54 | |
| *** wolverineav has quit IRC | 05:02 | |
| *** wolverineav has joined #openstack-neutron | 05:02 | |
| *** wolverineav has quit IRC | 05:08 | |
| *** dsneddon has joined #openstack-neutron | 05:09 | |
| *** dsneddon has quit IRC | 05:15 | |
| *** lpetrut has quit IRC | 05:15 | |
| *** ivve has quit IRC | 05:18 | |
| *** janki has quit IRC | 05:20 | |
| *** yamahata has joined #openstack-neutron | 05:23 | |
| *** dsneddon has joined #openstack-neutron | 05:29 | |
| *** wolverineav has joined #openstack-neutron | 05:34 | |
| *** dsneddon has quit IRC | 05:34 | |
| *** sridharg has joined #openstack-neutron | 05:39 | |
| *** janki has joined #openstack-neutron | 05:51 | |
| *** mosulica has joined #openstack-neutron | 06:01 | |
| *** dsneddon has joined #openstack-neutron | 06:01 | |
| *** gcheresh has joined #openstack-neutron | 06:05 | |
| *** dsneddon has quit IRC | 06:06 | |
| *** wolverineav has quit IRC | 06:07 | |
| *** iyamahat has quit IRC | 06:07 | |
| *** pcaruana has joined #openstack-neutron | 06:11 | |
| *** dsneddon has joined #openstack-neutron | 06:20 | |
| *** ivve has joined #openstack-neutron | 06:22 | |
| *** dsneddon has quit IRC | 06:25 | |
| *** ksambor has joined #openstack-neutron | 06:39 | |
| *** wolverineav has joined #openstack-neutron | 06:41 | |
| *** Luzi has joined #openstack-neutron | 06:44 | |
| *** slaweq has quit IRC | 06:45 | |
| *** sridharg has quit IRC | 06:50 | |
| *** sridharg has joined #openstack-neutron | 06:53 | |
| *** dsneddon has joined #openstack-neutron | 06:57 | |
| *** ivve has quit IRC | 06:58 | |
| *** lpetrut has joined #openstack-neutron | 07:02 | |
| *** ivve has joined #openstack-neutron | 07:03 | |
| *** dsneddon has quit IRC | 07:06 | |
| *** ccamposr has joined #openstack-neutron | 07:08 | |
| *** jlibosva has joined #openstack-neutron | 07:11 | |
| *** liuyulong_ has joined #openstack-neutron | 07:17 | |
| *** sridharg has quit IRC | 07:21 | |
| *** sridharg has joined #openstack-neutron | 07:23 | |
| *** rcernin has quit IRC | 07:24 | |
| *** hjensas has quit IRC | 07:25 | |
| *** cshen has joined #openstack-neutron | 07:26 | |
| *** matrohon has joined #openstack-neutron | 07:27 | |
| *** abaindur has quit IRC | 07:29 | |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: DNM Add some debug logs to linux utils module https://review.openstack.org/645032 | 07:32 |
| *** dsneddon has joined #openstack-neutron | 07:35 | |
| *** abaindur has joined #openstack-neutron | 07:39 | |
| *** slaweq has joined #openstack-neutron | 07:41 | |
| *** slaweq has quit IRC | 07:41 | |
| *** slaweq has joined #openstack-neutron | 07:41 | |
| *** luksky has joined #openstack-neutron | 07:46 | |
| *** dsneddon has quit IRC | 07:49 | |
| *** wolverineav has quit IRC | 07:54 | |
| *** ChosSimbaOne has joined #openstack-neutron | 07:59 | |
| *** lajoskatona has joined #openstack-neutron | 07:59 | |
| *** yamamoto has joined #openstack-neutron | 08:01 | |
| *** tesseract has joined #openstack-neutron | 08:02 | |
| *** hjensas has joined #openstack-neutron | 08:03 | |
| *** abaindur has quit IRC | 08:03 | |
| *** yamamoto has quit IRC | 08:05 | |
| *** liuyulong_ has quit IRC | 08:06 | |
| *** tssurya has quit IRC | 08:12 | |
| *** dsneddon has joined #openstack-neutron | 08:15 | |
| dulek | dalvarez: Hi! Can I discuss a networking-ovn issue with you? | 08:17 |
| dalvarez | dulek: o/ you can | 08:17 |
| *** openstackgerrit has quit IRC | 08:17 | |
| dulek | dalvarez: Ever noticed those: http://logs.openstack.org/62/608862/53/check/networking-ovn-tempest-dsvm-ovs-master/a67840c/controller/logs/screen-n-cpu.txt.gz#_Mar_19_13_42_55_705978 ? | 08:18 |
| dulek | dalvarez: It's coming from os-vif, we experience the same in Kuryr as we use os-vif for binding too. | 08:18 |
| dulek | dalvarez: It's because OVN's DevStack plugin configures OVS DB to bind to file socket, while os-vif (and Neutron too) config defaults point to 127.0.0.1:6440. | 08:20 |
| *** dsneddon has quit IRC | 08:20 | |
| dulek | dalvarez: So the solution would either be to bind that OVS DB to default address or put an option into Nova's (and Kuryr's) config that point to that file socket. | 08:21 |
| * dalvarez checks | 08:21 | |
| dulek | dalvarez: My question is - is there any reason OVN's DevStack plugin is using file socket instead of binding to port? | 08:21 |
| dalvarez | dulek: i wonder this is something we dont see in networking-ovn gate? | 08:22 |
| dulek | dalvarez: I've linked to networking-ovn gate results actually. | 08:23 |
| dalvarez | oh | 08:23 |
| dalvarez | sorry | 08:23 |
| *** gildub has joined #openstack-neutron | 08:23 | |
| dalvarez | dulek: sorry i fail to understand a bit, does this mean that instances fail to boot? | 08:23 |
| dulek | dalvarez: Just a random patch. I should probably say that this seems not affect anything functionally. | 08:24 |
| dulek | dalvarez: No, not at all, everything seems to work despite this traceback. | 08:24 |
| dalvarez | dulek: oh ok that's actually trying to create the integration bridge if it doesn't exist | 08:24 |
| *** tssurya has joined #openstack-neutron | 08:24 | |
| dalvarez | dulek: communicating to local OVSDB via unix socket is much faster and reliable | 08:25 |
| dulek | dalvarez: Ah, and that bridge exists, so it's redundant call? | 08:25 |
| numans | dulek, ovsdb-server can listen on both tcp socket and unix socket. And any client can choose either of the connection methods to talk to ovsdb-server. | 08:25 |
| dalvarez | dulek: so this is the main reason, and since it's local ovsdb, there's no real reason to use a tcp connection. On th eother hand, ovn-controller will create the integration bridge if it doesn't exist | 08:25 |
| dulek | numans: Oh, so I can just add another --remote 127.0.0.1:6640 to the command that starts ovs db in networking-ovn DevStack plugin? | 08:26 |
| dalvarez | dulek: we have some cases where we need to spawn a manager (in tripleo we do this) to connect to 127.0.0.1:6440 but this is because of the socket permissions | 08:26 |
| numans | dulek, that's right | 08:26 |
| dalvarez | numans: dulek i'd say it's best to tell os-vif to use the unix socket instead though, i wonder, how is os-vif plugging the interfaces later on? | 08:27 |
| numans | dalvarez, there is no harm though for networking-ovn devstack plugin to configure --remote tcp:127.0.0.1:6640 | 08:27 |
| numans | when it starts ovsdb-server for conf.db | 08:27 |
| dulek | dalvarez: For some reason it's only using explicit connection path in some commands. | 08:27 |
| dalvarez | numans: yes but i think it's best to use unix sockets | 08:27 |
| numans | dalvarez, that i agree | 08:27 |
| dalvarez | numans: dulek for example: http://logs.openstack.org/62/608862/53/check/networking-ovn-tempest-dsvm-ovs-master/a67840c/controller/logs/screen-n-cpu.txt.gz#_Mar_19_13_49_02_355776 | 08:27 |
| dulek | dalvarez: And some just do plain ovs-vsctl, which obviously works. | 08:28 |
| numans | dalvarez, and if os-vif has config option to set the ovsdb remote, then its better to configure there | 08:28 |
| dalvarez | looks like os-vif is failing on every other operation, how is it succeeding? :D | 08:28 |
| dalvarez | numans: ++ to that, i'd like that better | 08:28 |
| dalvarez | so see that trace | 08:28 |
| dalvarez | numans: dulek and after that 'Successfully unplugged vif VIFOpenVSwitch' how? :D | 08:28 |
| numans | dalvarez, may be networkign-ovn devstack plugin should set that ini config option | 08:28 |
| dulek | dalvarez: Magic? :D | 08:29 |
| dalvarez | maybe | 08:29 |
| dalvarez | dulek: numans i wonder... if tests pass that means that it's managing somehow to plug/unpug the ports into ovs | 08:29 |
| dalvarez | maybe trying both (tcp and unix socket?) | 08:29 |
| numans | dalvarez, ovs-vsctl will use unix path if --db is not specified | 08:30 |
| dalvarez | ahhhhh | 08:30 |
| dalvarez | numans++ <3 :D | 08:30 |
| dalvarez | ack then it's fine | 08:30 |
| numans | but here it loooks like --db is specified right ? | 08:30 |
| dalvarez | oh pok ok wait | 08:30 |
| dalvarez | numans: it's passing --db | 08:30 |
| numans | dalvarez, yeah | 08:30 |
| dulek | Yes, os-vif only passes --db sometimes. | 08:30 |
| dulek | For some commands. | 08:30 |
| dulek | It's weird, but that's how things are there. | 08:30 |
| dalvarez | yeah but you see that it logs 'successfully' | 08:31 |
| dulek | And yes, it's possible to set [os_vif_ovs]ovsdb_connection=<whatever> | 08:31 |
| dalvarez | so either it tries first with '--db' and then without it... or something's fishy cause it looks like it's certainly executing it ok | 08:31 |
| numans | dalvarez, easy fix is to pass --remote when starting ovsdb-server :) | 08:31 |
| dalvarez | jlibosva: may know :) | 08:32 |
| dalvarez | jlibosva: http://logs.openstack.org/62/608862/53/check/networking-ovn-tempest-dsvm-ovs-master/a67840c/controller/logs/screen-n-cpu.txt.gz#_Mar_19_13_49_02_355776 | 08:32 |
| * jlibosva | 08:32 | |
| dulek | numans: I'd actually prefer that. Otherwise to get rid of those you'd need to patch every project that uses networking-ovn and binds using os-vif. | 08:32 |
| dalvarez | jlibosva: and right after that error: Successfully unplugged vif VIFOpenVSwitch | 08:32 |
| dulek | As os-vif takes config from whatever project is using it. | 08:32 |
| numans | dulek, agree there. dalvarez i don't see any harm | 08:32 |
| numans | and i don't think there will be any performance degradation because of that | 08:33 |
| dulek | This means that in your plugin you need to put it in nova.conf. In Kuryr we'd need to put it in kuryr.conf and e.g. Zun would need to put it in zun.conf. | 08:33 |
| dalvarez | numans: dulek we can do that but imho if os_vif will read the config from the project, why not tell os-vif in networking-ovn to use unix socket? | 08:33 |
| dalvarez | dulek: ok | 08:33 |
| *** gkadam has joined #openstack-neutron | 08:33 | |
| *** gkadam has quit IRC | 08:33 | |
| dulek | dalvarez: That's the point, networking-ovn doesn't know who's going to bind the ports, so config filename is unknown. :P | 08:34 |
| dalvarez | i think it'd be best to change os-vif to default to unix socket instead | 08:34 |
| dalvarez | dulek: numans i agree there's not gonna be any perf degradation as it's localhost but there's no need, it's faster and more reliable so why not doing it better? :P | 08:35 |
| *** slaweq_ has joined #openstack-neutron | 08:36 | |
| jlibosva | dalvarez: dulek so what is the question for me? also the vswitchd says the port was actually deleted | 08:36 |
| dalvarez | jlibosva: basically question would be: if os-vif said it failed to execute the command, why the port was actually deleted? | 08:36 |
| dulek | dalvarez: /var/run/openvswitch/db.sock or /usr/local/var/run/openvswitch/db.sock? | 08:36 |
| dulek | dalvarez: It's not that simple networking-ovn uses a custom patch. | 08:36 |
| numans | dalvarez, if it solves the problem for every one just by fixing it in networking-ovn devstack files, i would say let's go with it without much complication or much work | 08:36 |
| dulek | s/patch/path | 08:36 |
| dalvarez | numans: ack :) | 08:37 |
| dalvarez | dulek: is that custom? i actually sent a patch to change everything to /var ... wait | 08:37 |
| numans | dalvarez, for production envs, the deployment tool can handle it properly :) | 08:37 |
| dalvarez | numans: yeah | 08:37 |
| dulek | dalvarez: Maybe I have older networking-ovn, it's possible. | 08:37 |
| dalvarez | dulek: numans https://review.openstack.org/#/c/643252/ | 08:37 |
| jlibosva | dalvarez: my guess is that os-vif says it lost connection to db, so DB could get the request to delete and then client lost connection? | 08:38 |
| jlibosva | but DB performed the action | 08:38 |
| numans | dalvarez, dulek also os-vif should have a config option to get the remote path | 08:38 |
| dulek | jlibosva: Nah, the DB doesn't even listen on that port. | 08:38 |
| dalvarez | jlibosva: there's nobody listening on 6440 | 08:38 |
| numans | its actually 6640 :) | 08:38 |
| dalvarez | numans: dulek yeah that's my point :D | 08:38 |
| dalvarez | atleeeti | 08:38 |
| dalvarez | yeah ok sorry 6640 :D | 08:38 |
| *** slaweq has quit IRC | 08:38 | |
| dulek | Whoops, my mistake. ;) | 08:38 |
| dalvarez | numans: dulek would this solve the thingy? https://review.openstack.org/#/c/643252/ | 08:39 |
| jlibosva | so how does ovs-vsctl talk to ovsdb? | 08:39 |
| dalvarez | cause we can go ahead and give it a whirl | 08:39 |
| dulek | dalvarez, numans: Okay, I'll dig a bit more to answer dalvarez question about why does it report success after failing. | 08:39 |
| dulek | And then propose the patch to open the 6640 | 08:39 |
| dalvarez | jlibosva: yeah idk from that log, networking-ovn uses unix socket | 08:39 |
| numans | jlibosva, by default when you run ovs-vsctl, it uses the unix socket path | 08:39 |
| jlibosva | numans: dalvarez but that still means it connects to db, no matter what IPC it uses, right? | 08:39 |
| dulek | dalvarez: This would help having a viable default in os-vif. But note that neutron.conf does default to 127.0.0.1:6640 as well. | 08:40 |
| numans | jlibosva, user can always use --db=tcp:IP:6640 or --db=unix:/... to override the default path | 08:40 |
| dalvarez | jlibosva: numans dulek but the command fails and the port still got deleted, wonder if os-vif or the vsctl imp thingy is using a fallback without '--db' | 08:40 |
| dulek | dalvarez: I'll check that. | 08:40 |
| dalvarez | that could be but still the command exits with != 0 | 08:40 |
| numans | jlibosva, in this case, ovs-vsctl is invoked as "--db=tcp:127.0.0.1:6640" and ovsdb-server is not listening on this port | 08:40 |
| numans | so it would fail connecting to the ovsdb-server | 08:40 |
| dalvarez | dulek: yeah neutron spawns a manager on 6640 which btw i dont know why, maybe jlibosva knows | 08:41 |
| dalvarez | but i dont see the reason, we do it in tripleo in ovn because of an issue with the permissions between the socket and the connecting processes | 08:41 |
| numans | jlibosva, dalvarez is suspecting that os-vif may be invoking "ovs-vsctl" again without --db option and that's why it is able to create ports | 08:41 |
| dalvarez | but in devstack i can't see why | 08:41 |
| numans | dalvarez, am i right ? | 08:42 |
| *** jpena|off is now known as jpena | 08:45 | |
| dalvarez | numans: you are! cause the operations succeeded | 08:45 |
| dulek | At the moment stuff I see in os-vif code does not indicate that, but I'm still digging. | 08:45 |
| *** slaweq_ has quit IRC | 08:47 | |
| dulek | Okay, I see now. | 08:48 |
| dulek | dalvarez: https://github.com/openstack/os-vif/blob/0c6a21c06406ca942c92e39a4a2b3fab61fefb60/vif_plug_ovs/ovsdb/impl_vsctl.py#L121 | 08:48 |
| dulek | dalvarez: https://github.com/openstack/os-vif/blob/0c6a21c06406ca942c92e39a4a2b3fab61fefb60/vif_plug_ovs/ovsdb/impl_vsctl.py#L110-L111 | 08:48 |
| dulek | Without explicitly putting check_error=True any OVS errors are not raised, only logged. | 08:49 |
| jlibosva | dalvarez: I'm back | 08:49 |
| jlibosva | dalvarez: I think there is some sort of manager in ovsdbapp that might open 6640 | 08:49 |
| jlibosva | let me see | 08:49 |
| dalvarez | dulek: what do you mean? that it's not true that the port was deleted? | 08:49 |
| dulek | dalvarez: I'd say that, yes. | 08:50 |
| *** alisanhaji has joined #openstack-neutron | 08:50 | |
| dulek | Unless something else sees unused port and deletes it. | 08:50 |
| dulek | But that call to ovs-vsctl haven't reached it. | 08:50 |
| *** ralonsoh has joined #openstack-neutron | 08:51 | |
| dalvarez | dulek: then how things work? ie. how tempest works if the VMs are not really plugged into the bridge? | 08:51 |
| dulek | dalvarez: It's only deletion that fail, right? | 08:52 |
| jlibosva | dalvarez: ovs agent opens 6640 but you don't have that, right? | 08:52 |
| dalvarez | dulek: are you sure? why ? | 08:53 |
| dalvarez | jlibosva: no, that's why. My question was more about why ovs agent does it ? for ryu? cause to me it makes sense to use just unix socket | 08:53 |
| dalvarez | i dont see why using tcp | 08:53 |
| dulek | dalvarez: And creation of integration bridge. | 08:53 |
| dulek | dalvarez: So two things that aren't really critical - bridge is already there and deletion just leaves orphans. | 08:53 |
| jlibosva | dalvarez: aha, I don't know why. perhaps it was because of parity with ovs-vsctl ovsdb driver | 08:54 |
| dulek | dalvarez: Okay, I'll check what's so special about those two commands, okay? | 08:54 |
| dalvarez | jlibosva: might be, anyways :) | 08:54 |
| dalvarez | dulek: thanks! | 08:54 |
| *** dsneddon has joined #openstack-neutron | 08:55 | |
| jlibosva | dalvarez: also os-vif has set that db connection: http://logs.openstack.org/62/608862/53/check/networking-ovn-tempest-dsvm-ovs-master/a67840c/controller/logs/screen-n-cpu.txt.gz#_Mar_19_13_38_17_429207 | 08:57 |
| dalvarez | jlibosva: yeah that's why it looks like os-vif is attempting to use --db=...:6640 | 08:58 |
| dalvarez | bt just for some commands | 08:59 |
| dalvarez | anyways | 08:59 |
| dalvarez | thanks jlibosva++ | 08:59 |
| jlibosva | I'm looking at os-vif and it looks like they mix ovsdbapp with vsctl | 08:59 |
| jlibosva | perhaps ralonsoh know more :) | 09:00 |
| *** dsneddon has quit IRC | 09:00 | |
| *** matrohon has quit IRC | 09:00 | |
| *** whoami-rajat has quit IRC | 09:10 | |
| *** whoami-rajat has joined #openstack-neutron | 09:18 | |
| *** sridharg has quit IRC | 09:19 | |
| ralonsoh | jlibosva, os-vif switched to ovsdbapp in the last release | 09:19 |
| ralonsoh | jlibosva, you still can choose vsctl, but native implementation is the default one | 09:20 |
| *** iyamahat has joined #openstack-neutron | 09:20 | |
| jlibosva | ralonsoh: dalvarez so this is set explicitly? http://logs.openstack.org/62/608862/53/check/networking-ovn-tempest-dsvm-ovs-master/a67840c/controller/logs/screen-n-cpu.txt.gz#_Mar_19_13_38_17_429741 | 09:21 |
| ralonsoh | jlibosva, let me check... | 09:22 |
| ralonsoh | jlibosva, dalvarez: well, I was wrong: the default interface is vsctl. I'll ask Sean. When I implemented the native connection, the plan was to switch to this one | 09:25 |
| ralonsoh | https://github.com/openstack/os-vif/blob/master/vif_plug_ovs/ovs.py#L67 | 09:26 |
| *** SASAA has joined #openstack-neutron | 09:27 | |
| *** sridharg has joined #openstack-neutron | 09:27 | |
| *** dsneddon has joined #openstack-neutron | 09:28 | |
| *** dsneddon has quit IRC | 09:33 | |
| dulek | dalvarez: Okay, it seems that this if clause is returning false: https://github.com/openstack/os-vif/blob/1546d349b1656bc98286af47f1e2748b76743abc/vif_plug_ovs/ovs.py#L251-L253 | 09:40 |
| dulek | dalvarez: Therefore the only ovs-vsctl operations run are the port deletion and ensuring that br-int exists. | 09:41 |
| dulek | dalvarez: And both fail. | 09:41 |
| dalvarez | dulek: got it | 09:42 |
| dalvarez | so we need to fix it :9 | 09:43 |
| dalvarez | :) | 09:43 |
| dulek | dalvarez: Yeah, it seems so. numans, do you know where the PortProfile is created in Neutron? | 09:44 |
| dulek | I want to double check that port_profile.create_port is set to false. | 09:44 |
| numans | dulek, i have no idea :) | 09:44 |
| *** pcaruana has quit IRC | 09:45 | |
| dulek | Okay, the comment in that method states that it's not called. | 09:46 |
| *** pcaruana has joined #openstack-neutron | 09:46 | |
| dulek | I'm putting up the patch to DevStack plugin, let's see if it helps. | 09:47 |
| *** wolverineav has joined #openstack-neutron | 09:51 | |
| *** cshen has quit IRC | 09:57 | |
| *** dsneddon has joined #openstack-neutron | 09:57 | |
| *** SmearedBeard has joined #openstack-neutron | 09:58 | |
| *** dsneddon has quit IRC | 10:02 | |
| *** slaweq_ has joined #openstack-neutron | 10:03 | |
| *** doude has joined #openstack-neutron | 10:04 | |
| *** cheng1 has quit IRC | 10:05 | |
| *** zbr has quit IRC | 10:08 | |
| *** matrohon has joined #openstack-neutron | 10:09 | |
| *** zbr has joined #openstack-neutron | 10:09 | |
| *** Smeared_Beard has joined #openstack-neutron | 10:10 | |
| *** cshen has joined #openstack-neutron | 10:11 | |
| *** SmearedBeard has quit IRC | 10:12 | |
| *** zbr has quit IRC | 10:16 | |
| *** slaweq_ is now known as slaweq | 10:18 | |
| dulek | dalvarez, numans: https://review.openstack.org/#/c/645096/ | 10:19 |
| dulek | I'll ping you again once I made sure it gets rids of tracebacks. | 10:19 |
| numans | dulek, nice. | 10:20 |
| numans | dulek, we wouldn't be knowing the HOST_IP right ? | 10:20 |
| numans | with your patch, ovsdb-server will be listening on 6640 on the IPv4's configured | 10:21 |
| dulek | numans: Ah, it's better to do ptcp:6640:127.0.0.1? | 10:21 |
| * dulek not sure about the exact format. | 10:21 | |
| numans | dulek, the format is correct. | 10:21 |
| numans | dulek, if ptcp:6640:127.0.0.1 works then i would suggest that. | 10:22 |
| numans | dulek, i think it should be good enough. | 10:22 |
| numans | dulek, from the logs we say --db=tcp:127.0.0.1:6640 right ? | 10:22 |
| numans | s/say/saw | 10:23 |
| dulek | numans: stack@devstack ~]$ sudo ovsdb-server --remote=punix:$OVS_RUNDIR/db.sock --remote=ptcp:127.0.0.1:6640 /opt/stack/data/ovs/conf.db | 10:24 |
| dulek | 2019-03-20T21:24:09Z|00001|socket_util|ERR|127.0.0.1:6640: bad port number "127.0.0.1" | 10:24 |
| dulek | [stack@devstack ~]$ sudo ovsdb-server --remote=punix:$OVS_RUNDIR/db.sock --remote=ptcp:6640:127.0.0.1 /opt/stack/data/ovs/conf.db | 10:24 |
| dulek | 2019-03-20T21:24:37Z|00001|ovsdb_server|INFO|ovsdb-server (Open vSwitch) 2.11.90 | 10:24 |
| numans | dulek ptcp:6640:127.0.0.1 is the right format. | 10:24 |
| numans | dulek, and when a client wants to connect it should use - tcp:127.0.0.1:6640 | 10:25 |
| dulek | Yes, it seems so. | 10:25 |
| dulek | And yes, the latter is the default for ovsdb_connection in os_vif. | 10:25 |
| numans | so lets do that way in your patch. | 10:25 |
| *** wolverineav has quit IRC | 10:25 | |
| dulek | numans: I've sent the fix, again - let's see if it gets rid of the tracebacks. | 10:29 |
| numans | dulek, thanks | 10:30 |
| *** cheng1 has joined #openstack-neutron | 10:32 | |
| *** priteau has joined #openstack-neutron | 10:35 | |
| *** dsneddon has joined #openstack-neutron | 10:35 | |
| *** alisanhaji has quit IRC | 10:40 | |
| *** dsneddon has quit IRC | 10:40 | |
| *** nicolasbock has joined #openstack-neutron | 10:58 | |
| *** openstackgerrit has joined #openstack-neutron | 10:59 | |
| openstackgerrit | Chengqian Liu proposed openstack/neutron-fwaas master: Fix the verification method before creating and updating the firewall rule https://review.openstack.org/618485 | 10:59 |
| *** dsneddon has joined #openstack-neutron | 11:04 | |
| *** njohnston_ has joined #openstack-neutron | 11:06 | |
| *** dsneddon has quit IRC | 11:10 | |
| *** njohnston_ is now known as njohnston | 11:18 | |
| *** priteau has quit IRC | 11:19 | |
| *** ileixe has quit IRC | 11:20 | |
| *** wolverineav has joined #openstack-neutron | 11:23 | |
| *** whoami-rajat has quit IRC | 11:30 | |
| openstackgerrit | Nate Johnston proposed openstack/neutron master: Utilize bulk port creation ops in ml2 plugin https://review.openstack.org/624815 | 11:31 |
| *** whoami-rajat has joined #openstack-neutron | 11:34 | |
| *** zbr has joined #openstack-neutron | 11:39 | |
| *** wolverineav has quit IRC | 11:40 | |
| *** luksky has quit IRC | 11:41 | |
| *** dsneddon has joined #openstack-neutron | 11:46 | |
| *** dave-mccowan has joined #openstack-neutron | 11:50 | |
| *** dsneddon has quit IRC | 11:51 | |
| *** panda is now known as panda|lunch | 11:52 | |
| *** dave-mccowan has quit IRC | 11:53 | |
| *** cshen has quit IRC | 11:53 | |
| *** tbachman has quit IRC | 11:57 | |
| *** jpena is now known as jpena|lunch | 12:01 | |
| *** dsneddon has joined #openstack-neutron | 12:06 | |
| *** dsneddon has quit IRC | 12:12 | |
| *** cshen has joined #openstack-neutron | 12:17 | |
| *** luksky has joined #openstack-neutron | 12:20 | |
| *** markvoelker has quit IRC | 12:22 | |
| *** dsneddon has joined #openstack-neutron | 12:22 | |
| *** tbachman has joined #openstack-neutron | 12:25 | |
| *** dsneddon has quit IRC | 12:27 | |
| *** dsneddon has joined #openstack-neutron | 12:28 | |
| *** goldyfruit has quit IRC | 12:31 | |
| *** dsneddon has quit IRC | 12:32 | |
| *** mriedem has joined #openstack-neutron | 12:38 | |
| *** cheng1 has quit IRC | 12:41 | |
| *** yamamoto has joined #openstack-neutron | 12:42 | |
| *** yamamoto has quit IRC | 12:43 | |
| *** yamamoto has joined #openstack-neutron | 12:51 | |
| *** dsneddon has joined #openstack-neutron | 12:59 | |
| openstackgerrit | boden proposed openstack/neutron-lib master: rehome neutron.db.standard_attr https://review.openstack.org/644977 | 13:02 |
| *** boden has joined #openstack-neutron | 13:03 | |
| *** yamamoto has quit IRC | 13:05 | |
| *** yamamoto has joined #openstack-neutron | 13:05 | |
| *** yamamoto has quit IRC | 13:06 | |
| *** dsneddon has quit IRC | 13:06 | |
| *** panda|lunch is now known as panda|sick | 13:07 | |
| *** irclogbot_3 has quit IRC | 13:07 | |
| *** gildub has quit IRC | 13:08 | |
| *** irclogbot_2 has joined #openstack-neutron | 13:09 | |
| *** jpena|lunch is now known as jpena | 13:20 | |
| *** altlogbot_0 has quit IRC | 13:23 | |
| *** altlogbot_2 has joined #openstack-neutron | 13:24 | |
| dulek | dalvarez, numans: Okay, no more tracebacks in n-cpu logs in https://review.openstack.org/#/c/645096/2. | 13:25 |
| dulek | Seems like it's working fine. | 13:25 |
| numans | dulek, cool | 13:25 |
| dalvarez | dulek: thanks | 13:26 |
| *** janki has quit IRC | 13:28 | |
| *** yamamoto has joined #openstack-neutron | 13:30 | |
| dalvarez | dulek: approved, let's see if we can make os-vif to use '--db' consistently across all the commands as well | 13:34 |
| dulek | dalvarez: It is used consistently. | 13:34 |
| dalvarez | dulek: not for adding ports, right? | 13:34 |
| dulek | dalvarez: It is. I'll link the code again, just a sec. | 13:34 |
| dalvarez | dulek: just for delete and bridge creation if i understood you correctly | 13:34 |
| *** gaoyan has joined #openstack-neutron | 13:35 | |
| dalvarez | dulek: ops then sry i missunderstood, why were ports getting correctly created? | 13:35 |
| dulek | dalvarez: https://github.com/openstack/os-vif/blob/1546d349b1656bc98286af47f1e2748b76743abc/vif_plug_ovs/ovs.py#L229-L253 | 13:35 |
| dulek | dalvarez: So ensure_ovs_bridge is happening per each port plugging. | 13:35 |
| dulek | dalvarez: But it's failure is harmless, right? | 13:35 |
| dalvarez | dulek: but i mean | 13:36 |
| dalvarez | dulek: when the port is created, i guess the ovs-vsctl call is not using --db=tcp:... right? | 13:36 |
| *** lemko has joined #openstack-neutron | 13:36 | |
| dulek | dalvarez: There's no other calls to ovs-vsctl happening on os-vif in case of port binding, it seems. | 13:37 |
| *** wolverineav has joined #openstack-neutron | 13:37 | |
| dulek | dalvarez: os-vif is only called on port plugging, I assume. | 13:37 |
| dalvarez | dulek: so how the port is added to the bridge? | 13:37 |
| *** njohnston has quit IRC | 13:37 | |
| dulek | dalvarez: I'm not an expert, so I don't 100% understand this comment, but I suppose it explains it: https://github.com/openstack/os-vif/blob/1546d349b1656bc98286af47f1e2748b76743abc/vif_plug_ovs/ovs.py#L233-L246 | 13:38 |
| *** njohnston has joined #openstack-neutron | 13:38 | |
| *** dsneddon has joined #openstack-neutron | 13:38 | |
| dulek | dalvarez: Therefore _create_vif_port is not called. | 13:38 |
| *** altlogbot_2 has quit IRC | 13:39 | |
| *** whoami-rajat has quit IRC | 13:40 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron-specs master: Port binding event extended information for Nova https://review.openstack.org/645173 | 13:41 |
| *** altlogbot_3 has joined #openstack-neutron | 13:41 | |
| *** dsneddon has quit IRC | 13:43 | |
| mriedem | there are a lot of uncategorized neutron functional job failures http://status.openstack.org/elastic-recheck/data/integrated_gate.html - is there some big known issue that has been fixed? | 13:43 |
| *** irclogbot_2 has quit IRC | 13:45 | |
| mriedem | http://logs.openstack.org/92/642892/1/check/neutron-functional/cafc553/job-output.txt.gz#_2019-03-21_12_21_33_642769 | 13:45 |
| slaweq | mriedem: yes, we have one issue which should be fixed with https://review.openstack.org/#/c/644842/ | 13:45 |
| mriedem | 2019-03-21 12:21:33.642769 | ERROR! 'msg' is not a valid attribute for a Task | 13:45 |
| mriedem | ack thanks i'll put up an e-r query for that | 13:45 |
| *** irclogbot_1 has joined #openstack-neutron | 13:46 | |
| mloza | Hello, how can I modify the RBAC to allow a user with member role create an instance in an external network? | 13:47 |
| *** yamamoto has quit IRC | 13:48 | |
| *** yamamoto has joined #openstack-neutron | 13:50 | |
| *** jistr is now known as jistr|call | 13:52 | |
| *** hjensas has quit IRC | 13:52 | |
| *** lajoskatona has quit IRC | 13:54 | |
| *** jistr|call is now known as jistr | 14:00 | |
| *** mlavalle has joined #openstack-neutron | 14:05 | |
| mriedem | mlavalle: can we get https://review.openstack.org/#/c/640390 merged and deal with nits in a follow up? it's holding up nova docs and release notes b/c of depends-on | 14:05 |
| mriedem | looks like bence aleady posted https://review.openstack.org/#/c/645104/ | 14:06 |
| mlavalle | mriedem: done | 14:06 |
| *** lpetrut has quit IRC | 14:06 | |
| *** lpetrut has joined #openstack-neutron | 14:07 | |
| mriedem | thanks | 14:07 |
| mlavalle | mriedem: thank you | 14:07 |
| mlavalle | slaweq: still waiting for https://review.openstack.org/#/c/644842, right? | 14:11 |
| *** dsneddon has joined #openstack-neutron | 14:11 | |
| slaweq | mlavalle: yes | 14:12 |
| slaweq | unfortunatelly :/ | 14:12 |
| mlavalle | slaweq: cool, so hold rechecks | 14:12 |
| *** wolverineav has quit IRC | 14:12 | |
| slaweq | and looking at http://zuul.openstack.org/status it again failed in gate :/ | 14:12 |
| slaweq | so will be another recheck for it :( | 14:12 |
| mlavalle | slaweq: thanks for following up on that. when you need to leave, if needed, let me know and I'll continue pushing it foward | 14:13 |
| *** dsneddon has quit IRC | 14:16 | |
| slaweq | mlavalle: sure, thx | 14:17 |
| slaweq | I'm now still around and looking at it | 14:17 |
| mlavalle | ok | 14:17 |
| slaweq | I know it's urgent as it blocks everything else now so we can't release RC1 without it :/ | 14:17 |
| *** whoami-rajat has joined #openstack-neutron | 14:23 | |
| *** gaoyan has quit IRC | 14:25 | |
| *** hongbin has joined #openstack-neutron | 14:29 | |
| *** altlogbot_3 has quit IRC | 14:35 | |
| *** yamamoto has quit IRC | 14:36 | |
| *** altlogbot_1 has joined #openstack-neutron | 14:36 | |
| *** yamamoto has joined #openstack-neutron | 14:37 | |
| *** irclogbot_1 has quit IRC | 14:38 | |
| *** irclogbot_2 has joined #openstack-neutron | 14:39 | |
| openstackgerrit | Nate Johnston proposed openstack/neutron master: [DNM] CI test https://review.openstack.org/644594 | 14:42 |
| *** yamamoto has quit IRC | 14:42 | |
| *** hjensas has joined #openstack-neutron | 14:44 | |
| *** dsneddon has joined #openstack-neutron | 14:47 | |
| frickler | mlavalle: slaweq: instead of doing umpteen rechecks, how about force-merging this? seems it is an obvious fix only affecting zuul config and not any code. I'd be happy to do that for you if you want | 14:49 |
| slaweq | frickler: I'm fine with that | 14:49 |
| mlavalle | frickler: thanks. let's do it | 14:49 |
| slaweq | failures which prevent this patch to be merged weren't for sure related to it | 14:49 |
| slaweq | mlavalle++ | 14:49 |
| slaweq | thx frickler for help :) | 14:49 |
| *** gcheresh has quit IRC | 14:50 | |
| *** matrohon has quit IRC | 14:50 | |
| *** gaoyan has joined #openstack-neutron | 14:50 | |
| mlavalle | njohnston: please see conversation with frickler^^^^. do you need to rebase again? | 14:50 |
| *** panda|sick is now known as panda|drappt | 14:50 | |
| mlavalle | https://review.openstack.org/#/c/624815 | 14:51 |
| *** gaoyan has quit IRC | 14:51 | |
| frickler | mlavalle: we are talking about merging https://review.openstack.org/644842 , just to avoid any misinterpretation, correct? | 14:52 |
| *** dsneddon has quit IRC | 14:52 | |
| mlavalle | frickler: yes | 14:53 |
| mlavalle | njohnston: nevermind | 14:53 |
| slaweq | frickler: correct | 14:55 |
| openstackgerrit | Merged openstack/neutron master: Remove "msg" from tasks in ansible roles https://review.openstack.org/644842 | 14:59 |
| *** dsneddon has joined #openstack-neutron | 15:00 | |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Try to enable dnsmasq process several times https://review.openstack.org/643732 | 15:01 |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Fix neutron-status upgrade check CLI tool https://review.openstack.org/643485 | 15:01 |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Switch neutron-functional-with-wsgi job to zuul v3 syntax https://review.openstack.org/643702 | 15:01 |
| mlavalle | slaweq: would you push https://review.openstack.org/#/c/613820/? | 15:05 |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Define qos-rules-alias extension https://review.openstack.org/613820 | 15:06 |
| slaweq | mlavalle: done | 15:06 |
| frickler | mlavalle: slaweq: done. per the discussion in -infra, next time you are experiencing a similar issue, you could also consider amending your fix-patch with disabling/disvoting all the unstable jobs, allowing it to merge more easily. then in a second patch revert the latter part, possibly combined with fixes for the unstable jobs | 15:06 |
| mlavalle | slaweq: thanks! | 15:06 |
| slaweq | I rebased it to get this "msg" fix in | 15:06 |
| mlavalle | slaweq: yeap, thanks | 15:07 |
| *** hjensas has quit IRC | 15:07 | |
| mlavalle | frickler: noted. Thanks for the advice :-) | 15:08 |
| *** wolverineav has joined #openstack-neutron | 15:08 | |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: More accurate agent restart state transfer https://review.openstack.org/640797 | 15:12 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Change default local ovs connection timeout https://review.openstack.org/638645 | 15:13 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Adds revision description for network_segment_range https://review.openstack.org/643882 | 15:13 |
| *** mosulica has quit IRC | 15:14 | |
| *** liuyulong|away is now known as liuyulong | 15:17 | |
| liuyulong | https://review.openstack.org/640797, slaweq, mlavalle, boden, hi, just rebase the patch, please help add the +W, thanks. | 15:18 |
| mlavalle | liuyulong: Looking at njohnston's patch now. If nobody beats me to it, I'll get to yours afterwards | 15:20 |
| njohnston | mlavalle: Thanks! | 15:20 |
| liuyulong | njohnston, OK, hi, maybe you can add your +w back, : ) | 15:20 |
| *** sapd1_x has joined #openstack-neutron | 15:21 | |
| *** Luzi has quit IRC | 15:25 | |
| f0o | boden: Hi, heard anything new yet? I dont relly need a fix for it, just a pointer where to look at it to path it myself. I'm not very versed with the magic callbacks in the stacktrace and they're rather hard to find too | 15:30 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Ignore first local port update notification https://review.openstack.org/644613 | 15:31 |
| *** mattw4 has joined #openstack-neutron | 15:36 | |
| openstackgerrit | Bence Romsics proposed openstack/neutron master: FUP: Unit tests for https://review.openstack.org/636970 https://review.openstack.org/645210 | 15:40 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: Change provider network segmentation ID https://review.openstack.org/633165 | 15:41 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: Decouple local VLAN internal management and OVS enforcement https://review.openstack.org/644385 | 15:41 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: Change provider network segmentation ID in OVS agent https://review.openstack.org/632984 | 15:41 |
| *** wolverineav has quit IRC | 15:42 | |
| *** ivve has quit IRC | 15:42 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: Increase timeouts for OVSDB in functional tests https://review.openstack.org/642721 | 15:42 |
| mlavalle | haleyb, slaweq: I'm ok with https://review.openstack.org/#/c/624815. Please take another look when you have a chance | 15:43 |
| mlavalle | njohnston: ^^^^ | 15:43 |
| *** mattw4 has quit IRC | 15:44 | |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Random IP for concurrent create pf and update_port https://review.openstack.org/644278 | 15:45 |
| openstackgerrit | Bence Romsics proposed openstack/neutron master: Reject min-bw rule operations on non-physnet networks/ports https://review.openstack.org/641712 | 15:47 |
| openstackgerrit | Bence Romsics proposed openstack/neutron master: FUP: Unit tests for https://review.openstack.org/636970 https://review.openstack.org/645210 | 15:47 |
| *** mattw4 has joined #openstack-neutron | 15:48 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: Remove conntrack rule when FIP is deleted https://review.openstack.org/641747 | 15:50 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Change ovs-agent iteration log level to INFO https://review.openstack.org/638641 | 15:52 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Add test case to duplicated service plugin https://review.openstack.org/638447 | 15:52 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Divide-and-conquer security group beasts https://review.openstack.org/638642 | 15:53 |
| openstackgerrit | LIU Yulong proposed openstack/neutron master: Do not call update_device_list in large sets https://review.openstack.org/638646 | 15:53 |
| *** cshen has quit IRC | 15:53 | |
| *** hjensas has joined #openstack-neutron | 15:58 | |
| openstackgerrit | Merged openstack/neutron master: Networking guide: Guaranteed Minimum Bandwidth https://review.openstack.org/640390 | 15:59 |
| *** dsneddon has quit IRC | 16:08 | |
| openstackgerrit | Merged openstack/neutron master: FUP Networking guide: Guaranteed Minimum Bandwidth https://review.openstack.org/645104 | 16:09 |
| openstackgerrit | Merged openstack/networking-ovn master: DevStack: Make ovsdb-server listen on 6640 https://review.openstack.org/645096 | 16:10 |
| boden | f0o sorry your question is in regards to? | 16:12 |
| boden | f0o the vmware-nsx bug? | 16:12 |
| *** goldyfruit has joined #openstack-neutron | 16:13 | |
| *** wwriverrat has quit IRC | 16:13 | |
| *** sapd1_x has quit IRC | 16:16 | |
| *** dsneddon has joined #openstack-neutron | 16:17 | |
| f0o | boden: yeah the vmware-nsx issue with the execption about router_distributed | 16:19 |
| *** imacdonn has joined #openstack-neutron | 16:19 | |
| *** mattw4 has quit IRC | 16:21 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: Check master/backup router status in DVR functional tests https://review.openstack.org/645225 | 16:22 |
| *** cmart has joined #openstack-neutron | 16:23 | |
| *** dsneddon has quit IRC | 16:23 | |
| *** hjensas has quit IRC | 16:24 | |
| *** lpetrut has quit IRC | 16:32 | |
| *** gildub has joined #openstack-neutron | 16:34 | |
| *** tssurya has quit IRC | 16:34 | |
| *** cshen has joined #openstack-neutron | 16:37 | |
| *** cshen has quit IRC | 16:42 | |
| *** mvkr has quit IRC | 16:48 | |
| *** rubasov has quit IRC | 16:49 | |
| boden | f0o I think the dev team is trying to get things wrapped up for stein as of recent... let me double back with them again to ensure it's on the radar | 16:54 |
| *** ivve has joined #openstack-neutron | 16:54 | |
| *** rpittau is now known as rpittau|afk | 16:55 | |
| f0o | boden: thanks and understandable, like I said just a hint where to look for would be enough for me to start tackling the issue heh | 16:55 |
| *** dsneddon has joined #openstack-neutron | 17:00 | |
| *** luksky has quit IRC | 17:02 | |
| *** tesseract has quit IRC | 17:02 | |
| *** yamahata has quit IRC | 17:04 | |
| *** iyamahat has quit IRC | 17:04 | |
| boden | f0o I left a comment in the bug | 17:04 |
| boden | we can discuss there... I may end up at a dead end, but can try to help | 17:04 |
| *** dsneddon has quit IRC | 17:05 | |
| f0o | thanks :) | 17:05 |
| f0o | boden: I have tried adding the option to the [DEFAULT] section but it didnt change anything | 17:05 |
| boden | f0o in neutron.conf correct?? and then you restarted neutron? | 17:05 |
| f0o | yep | 17:06 |
| *** igordc has joined #openstack-neutron | 17:06 | |
| f0o | boden in case you wonder, we only have 1 AZ in neutron which is `default` - since the last process in the stack was about processing AZ's | 17:07 |
| *** iyamahat has joined #openstack-neutron | 17:16 | |
| *** mriedem is now known as mriedem_grooming | 17:16 | |
| *** mvkr has joined #openstack-neutron | 17:16 | |
| openstackgerrit | Swaminathan Vasudevan proposed openstack/neutron master: Packets getting lost during SNAT with too many connections https://review.openstack.org/636473 | 17:20 |
| boden | f0o so I'm kind of guessing in the dark here, but it seems the issue may be that the router_distributed is not registred for the conf when the vmware-nsx plugin loads.. is it possible for you to try a hack and see if something works? | 17:20 |
| f0o | sure thing, throw me a patch :) | 17:20 |
| boden | f0o edit vmware_nsx.plugins.nsx_v.plugin.py and atop the file before the other imports add the following line | 17:22 |
| boden | from neutron.db import l3_dvr_db | 17:22 |
| boden | f0o then save and try restarting neutron | 17:22 |
| f0o | will do | 17:23 |
| boden | f0o in case your wondering by importing that it should register the conf opts as per https://github.com/openstack/neutron/blob/a388701ddfe628e9a5bd16a78422164799b11ef8/neutron/db/l3_dvr_db.py#L56 | 17:24 |
| *** dsneddon has joined #openstack-neutron | 17:24 | |
| f0o | yeah I stumbled across that file but thought it would be included elsewhere | 17:26 |
| f0o | I'm just pushing the change into the build-pipeline (devops is great until you need something fast) | 17:26 |
| *** mattw4 has joined #openstack-neutron | 17:32 | |
| *** gcheresh has joined #openstack-neutron | 17:32 | |
| *** yamahata has joined #openstack-neutron | 17:34 | |
| boden | f0o okay.. it's not clear to me if importing it into that file is early enough in the process, but I guess we'll see | 17:34 |
| boden | also I'm not sure this is anything vmware specific, but I guess no proof otherwise ATM | 17:35 |
| *** ccamposr has quit IRC | 17:36 | |
| f0o | I got my fingers crossed :D | 17:39 |
| *** gmann is now known as gmann_afk | 17:43 | |
| *** rubasov has joined #openstack-neutron | 17:43 | |
| f0o | hrm something else blew up now | 17:45 |
| f0o | AttributeError: 'module' object has no attribute 'NSX_EXT_PATH' | 17:45 |
| f0o | just gonna check that I didnt mess up the patch | 17:45 |
| *** mvkr has quit IRC | 17:46 | |
| f0o | neutron_extensions.append_api_extensions_path([vmware_nsx.NSX_EXT_PATH]) in vmware_nsx.plugins.nsx_v.plugin.py (line 268/267) is now failing | 17:47 |
| *** cshen has joined #openstack-neutron | 17:48 | |
| f0o | I think I see what's happening | 17:48 |
| *** cshen has quit IRC | 17:52 | |
| boden | f0o not enough info for me to help... got a stack trace or somethign? | 17:53 |
| f0o | so NSX_EXT_PATH is defined in vmware_nsx/__init__.py right? - But when I try using vmware_nsx.plugin.NsxVPlugin as plugin (which worked in RDO) causes RHEL to just fail. So I need to use the full path (vmware_nsx.plugins.nsx_v.plugin.NsxVPluginV2) instead. However this apparently doesnt make the vmware_nsx/__init__.py load thus a lot of things going weird | 17:53 |
| f0o | just a blind guess | 17:54 |
| *** ivve has quit IRC | 17:54 | |
| boden | f0o: is there a stack trace I can see? maybe on openstack pastebin or something? | 17:56 |
| f0o | it comes as JSON, is that fine? | 17:56 |
| boden | f0o I can try | 17:57 |
| f0o | http://paste.openstack.org/show/748195/ | 17:58 |
| *** gcheresh has quit IRC | 17:58 | |
| f0o | boden: http://paste.openstack.org/show/748196/ (nicer paste) | 18:00 |
| boden | f0o so this is in RDO? | 18:02 |
| *** ivve has joined #openstack-neutron | 18:02 | |
| f0o | this is RHEL's OpenStack mumboyumbo | 18:02 |
| f0o | RDO works fine | 18:02 |
| f0o | sadly we're moving from RDO to RHEL because of $reasons :| | 18:02 |
| boden | f0o: even more reason to wait for the vmware-nsx plugin team; I'm not familiar with either and it's not clear why it would work in 1 and not the other if the code/config is the same | 18:04 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron-lib master: Add new vif_details parameters https://review.openstack.org/645288 | 18:05 |
| f0o | boden: I appreciate the attempts tho, I will play around with defining the oslo-config option in multiple parts of the driver to see if that just solves it by random | 18:05 |
| f0o | I'll keep the bug-report updated | 18:05 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: [DNM] Log OVS lib min bw rule https://review.openstack.org/642000 | 18:05 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron master: [DNM] Test test_reschedule_network_on_new_agent https://review.openstack.org/643079 | 18:06 |
| *** gildub has quit IRC | 18:16 | |
| *** jpena is now known as jpena|off | 18:22 | |
| *** whoami-rajat has quit IRC | 18:23 | |
| *** yamamoto has joined #openstack-neutron | 18:27 | |
| *** ralonsoh has quit IRC | 18:28 | |
| *** yamamoto has quit IRC | 18:31 | |
| *** cshen has joined #openstack-neutron | 18:32 | |
| *** ivve has quit IRC | 18:36 | |
| *** cshen has quit IRC | 18:37 | |
| *** gmann_afk is now known as gmann | 18:40 | |
| mlavalle | haleyb: so 55 revisions is your limit? good to know :-) | 18:41 |
| openstackgerrit | boden proposed openstack/neutron-lib master: rehome neutron.objects.extensions.standardattributes https://review.openstack.org/645300 | 18:46 |
| *** Smeared_Beard has quit IRC | 18:51 | |
| haleyb | mlavalle: i've seen worse i guess, but that is a lot! | 18:52 |
| *** mriedem_grooming is now known as mriedem | 18:53 | |
| *** itlinux has quit IRC | 18:55 | |
| *** abaindur has joined #openstack-neutron | 18:57 | |
| *** abaindur has quit IRC | 18:58 | |
| *** abaindur has joined #openstack-neutron | 18:58 | |
| *** awalende has joined #openstack-neutron | 19:11 | |
| *** mvkr has joined #openstack-neutron | 19:13 | |
| *** whoami-rajat has joined #openstack-neutron | 19:26 | |
| *** hjensas has joined #openstack-neutron | 19:28 | |
| gmann | yamahata: i replied on your query, can you recheck this now - https://review.openstack.org/#/c/639990/ | 19:28 |
| gmann | mlavalle: amotoki need another +2 on this - https://review.openstack.org/#/c/639361/11 | 19:28 |
| *** luksky has joined #openstack-neutron | 19:29 | |
| *** matrohon has joined #openstack-neutron | 19:37 | |
| haleyb | gmann: question on that review - why didn't it change to ubuntu-bionic-2-node ? | 19:46 |
| *** sridharg has quit IRC | 19:46 | |
| gmann | haleyb: on 639361 ? | 19:46 |
| haleyb | gmann: yes | 19:47 |
| *** rkukura has quit IRC | 19:47 | |
| haleyb | they are multinode jobs | 19:47 |
| *** gildub has joined #openstack-neutron | 19:47 | |
| gmann | haleyb: that is going to be taken care by base job with branch variant | 19:47 |
| gmann | http://git.openstack.org/cgit/openstack-infra/openstack-zuul-jobs/tree/zuul.d/jobs.yaml#n1098 | 19:48 |
| gmann | so we no more need to override the nodeset in derived job. previously there was no nodeset in base job | 19:48 |
| haleyb | gmann: ack, thanks, i'll +2 it | 19:48 |
| gmann | haleyb: thanks | 19:49 |
| mlavalle | gmann: thanks for the clarification | 19:52 |
| gmann | np! | 19:55 |
| f0o | boden: I think you just fixed it... I introduced a bunch of false-errors because the build-pipeline was yolo'd but after copying the content of l3_dvr_db to register the config options it's now creating routers. I will try with the import tomorrow to make it look nicer. But it's looking bliss so far! | 20:01 |
| boden | f0o interesting... once you get done and have time, do you mind updating the bug? | 20:02 |
| f0o | Ofc! | 20:03 |
| *** jlibosva has quit IRC | 20:07 | |
| *** rkukura has joined #openstack-neutron | 20:10 | |
| *** gildub has quit IRC | 20:31 | |
| *** armax has joined #openstack-neutron | 20:32 | |
| *** igordc has quit IRC | 20:32 | |
| *** cshen has joined #openstack-neutron | 20:33 | |
| *** cshen has quit IRC | 20:38 | |
| mlavalle | haleyb: I have to give my +1 to https://review.openstack.org/#/c/644330. are we ok releasing as rc1? Please note that this is not the latest commit in the repo: https://github.com/openstack/networking-ovn/commits/master. There are 5 new commits | 20:51 |
| *** wwriverrat has joined #openstack-neutron | 20:59 | |
| *** armax has quit IRC | 21:00 | |
| mlavalle | dalvarez: question is also for you^^^^ | 21:10 |
| mlavalle | tidwellr: are we ok to release as rc1: https://review.openstack.org/#/c/644333? the hash is the same as the latest commit in the repo | 21:11 |
| *** cmart has quit IRC | 21:12 | |
| tidwellr | mlavalle: that is correct | 21:14 |
| mlavalle | tidwellr: RC1 it is then. Thanks! | 21:14 |
| tidwellr | mlavalle: let's go for it | 21:14 |
| openstackgerrit | OpenStack Release Bot proposed openstack/networking-bagpipe master: Update master for stable/stein https://review.openstack.org/645356 | 21:15 |
| *** rkukura has quit IRC | 21:16 | |
| *** armax has joined #openstack-neutron | 21:17 | |
| openstackgerrit | Merged openstack/neutron master: Utilize bulk port creation ops in ml2 plugin https://review.openstack.org/624815 | 21:20 |
| njohnston | ^^ \o/ | 21:20 |
| openstackgerrit | OpenStack Release Bot proposed openstack/networking-bgpvpn master: Update master for stable/stein https://review.openstack.org/645360 | 21:20 |
| *** SASAA has quit IRC | 21:27 | |
| openstackgerrit | Merged openstack/networking-odl master: Use latest networking-sfc release https://review.openstack.org/640824 | 21:28 |
| *** matrohon has quit IRC | 21:29 | |
| *** pcaruana has quit IRC | 21:33 | |
| mlavalle | njohnston: thanks for your hard work! | 21:34 |
| openstackgerrit | Merged openstack/neutron master: Define qos-rules-alias extension https://review.openstack.org/613820 | 21:38 |
| openstackgerrit | Merged openstack/neutron master: Fix neutron-status upgrade check CLI tool https://review.openstack.org/643485 | 21:38 |
| openstackgerrit | OpenStack Release Bot proposed openstack/neutron-dynamic-routing master: Update master for stable/stein https://review.openstack.org/645370 | 21:40 |
| *** idlemind has joined #openstack-neutron | 21:44 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/networking-midonet master: Update master for stable/stein https://review.openstack.org/645378 | 21:45 |
| *** whoami-rajat has quit IRC | 21:45 | |
| *** armax has quit IRC | 21:45 | |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Migrate legacy jobs to Ubuntu Bionic https://review.openstack.org/639361 | 21:48 |
| mlavalle | njohnston: you still around? | 21:49 |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Add check for external_network_bridge in neutron-status https://review.openstack.org/643486 | 21:50 |
| *** rkukura has joined #openstack-neutron | 21:51 | |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron-fwaas master: Switch neutron-fwaas-fullstack job to zuulv3 syntax https://review.openstack.org/644526 | 21:54 |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron master: Dynamically increase DHCP process queue green pool size https://review.openstack.org/641490 | 21:55 |
| *** njohnston_ has joined #openstack-neutron | 21:59 | |
| *** njohnston has quit IRC | 21:59 | |
| openstackgerrit | Merged openstack/networking-odl master: Add bgpvpn-vni as a supported extension for ODL BGPVPN driver https://review.openstack.org/614428 | 22:00 |
| mlavalle | manjeets: this is the latest patch that merged in odl: https://review.openstack.org/#/c/614428. I am going to update the rc-1 release patch with it and we will go with that. ok? | 22:05 |
| mlavalle | slaweq: let's talk here | 22:05 |
| *** irclogbot_2 has quit IRC | 22:05 | |
| slaweq | mlavalle: sure | 22:06 |
| openstackgerrit | Matt Welch proposed openstack/neutron master: Enforce router admin state before distributed https://review.openstack.org/625134 | 22:06 |
| *** cmart has joined #openstack-neutron | 22:07 | |
| manjeets | mlavalle, looks good and thanks for reviews :) | 22:07 |
| mlavalle | manjeets: yw :-) | 22:08 |
| slaweq | mlavalle: I think I'm fine with https://review.openstack.org/#/c/641712/ | 22:08 |
| openstackgerrit | OpenStack Release Bot proposed openstack/networking-sfc master: Update master for stable/stein https://review.openstack.org/645387 | 22:09 |
| slaweq | I don't think we will break API backward compatibility with stable branches | 22:09 |
| *** armax has joined #openstack-neutron | 22:09 | |
| slaweq | so in such case I don't think that API extension is necessary | 22:09 |
| mlavalle | slaweq: ok please +2 it and then I'll approve it. we will go with this commit for out rc-1 | 22:09 |
| mlavalle | our^^^^ | 22:10 |
| mlavalle | Thanks! | 22:10 |
| slaweq | I just +2 both https://review.openstack.org/#/c/644847/1 and https://review.openstack.org/#/c/641712/ | 22:10 |
| *** ivve has joined #openstack-neutron | 22:11 | |
| mlavalle | slaweq: ok, I just did the same. thanks! | 22:13 |
| slaweq | mlavalle: thx, I'm going to sleep now. Good night! | 22:13 |
| slaweq | and see You tomorrow | 22:13 |
| *** cmart has quit IRC | 22:13 | |
| mlavalle | slaweq: good night. You need your beauty sleep | 22:13 |
| *** rcernin has joined #openstack-neutron | 22:19 | |
| *** armax has quit IRC | 22:20 | |
| mlavalle | johnsom: we don't need this, right https://review.openstack.org/#/c/644337/? | 22:21 |
| *** cmart has joined #openstack-neutron | 22:22 | |
| *** boden has quit IRC | 22:24 | |
| johnsom | mlavalle I am managing that patch/release | 22:26 |
| mlavalle | johnsom: cool, thanks! | 22:27 |
| *** armax has joined #openstack-neutron | 22:27 | |
| *** armax has quit IRC | 22:32 | |
| *** cshen has joined #openstack-neutron | 22:33 | |
| *** armax has joined #openstack-neutron | 22:35 | |
| *** cshen has quit IRC | 22:37 | |
| *** armax has quit IRC | 22:38 | |
| *** cmart has quit IRC | 22:38 | |
| *** mriedem is now known as mriedem_afk | 22:50 | |
| openstackgerrit | Merged openstack/neutron-fwaas master: Fix the verification method before creating and updating the firewall rule https://review.openstack.org/618485 | 22:51 |
| *** cmart has joined #openstack-neutron | 22:54 | |
| *** tkajinam has joined #openstack-neutron | 22:56 | |
| *** hongbin has quit IRC | 22:57 | |
| *** igordc has joined #openstack-neutron | 23:04 | |
| *** armax has joined #openstack-neutron | 23:09 | |
| *** mattw4 has quit IRC | 23:10 | |
| mlavalle | liuyulong: ping | 23:26 |
| *** armax has quit IRC | 23:26 | |
| *** ivve has quit IRC | 23:27 | |
| mlavalle | amotoki: you around? | 23:30 |
| *** awalende has quit IRC | 23:30 | |
| *** luksky has quit IRC | 23:32 | |
| *** goldyfruit has quit IRC | 23:35 | |
| *** wolverineav has joined #openstack-neutron | 23:40 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/networking-odl master: Update master for stable/stein https://review.openstack.org/645403 | 23:41 |
| *** wolverineav has quit IRC | 23:45 | |
| liuyulong | mlavalle, pong, hi | 23:57 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!