| johnsom | kong rm_work So I commented. Very tired after being on a customer call for the second straight day. | 00:00 |
|---|---|---|
| johnsom | I didn't change the amphora API, so really it is about making sure you change your config file at the same time you deploy your new amphora image. | 00:01 |
| johnsom | All of my code changes were inside the agent code, none on the controller | 00:01 |
| johnsom | How the agent in the amp interprets those API calls is up to the agent in the image. | 00:02 |
| *** cody-somerville has joined #openstack-lbaas | 00:27 | |
| *** cody-somerville has quit IRC | 00:47 | |
| *** cody-somerville has joined #openstack-lbaas | 00:48 | |
| *** yuanying has quit IRC | 00:55 | |
| *** BjoernT has joined #openstack-lbaas | 00:57 | |
| *** kevo has quit IRC | 00:59 | |
| *** cody-somerville has quit IRC | 01:06 | |
| *** amotoki has quit IRC | 01:08 | |
| *** cody-somerville has joined #openstack-lbaas | 01:10 | |
| *** cody-somerville has quit IRC | 01:16 | |
| *** ducttape_ has joined #openstack-lbaas | 01:21 | |
| *** cody-somerville has joined #openstack-lbaas | 01:23 | |
| *** prabampm has joined #openstack-lbaas | 01:24 | |
| *** woodster_ has quit IRC | 01:28 | |
| *** prabampm has quit IRC | 01:30 | |
| *** cody-somerville has quit IRC | 01:39 | |
| *** ducttape_ has quit IRC | 01:42 | |
| *** bana_k has quit IRC | 01:43 | |
| *** amotoki has joined #openstack-lbaas | 01:48 | |
| *** meghal has quit IRC | 01:56 | |
| *** amotoki has quit IRC | 01:56 | |
| *** yuanying has joined #openstack-lbaas | 01:59 | |
| *** amotoki has joined #openstack-lbaas | 01:59 | |
| *** amotoki has quit IRC | 02:09 | |
| *** amotoki has joined #openstack-lbaas | 02:18 | |
| openstackgerrit | Merged openstack/octavia: Add env variables for memory and vcpu in the Vagrantfile https://review.openstack.org/319870 | 02:48 |
| *** mixos has joined #openstack-lbaas | 02:48 | |
| *** fnaval has quit IRC | 02:55 | |
| *** fnaval has joined #openstack-lbaas | 03:06 | |
| *** yuanying has quit IRC | 03:14 | |
| *** BjoernT has quit IRC | 03:14 | |
| *** cody-somerville has joined #openstack-lbaas | 03:40 | |
| *** yamamoto_ has quit IRC | 03:43 | |
| *** cody-somerville_ has joined #openstack-lbaas | 03:43 | |
| *** cody-somerville has quit IRC | 03:46 | |
| *** cody-somerville_ has quit IRC | 03:51 | |
| *** meghal has joined #openstack-lbaas | 03:54 | |
| *** links has joined #openstack-lbaas | 04:01 | |
| *** cody-somerville_ has joined #openstack-lbaas | 04:09 | |
| *** kobis has joined #openstack-lbaas | 04:17 | |
| *** kobis has quit IRC | 04:18 | |
| *** cody-somerville_ has quit IRC | 04:20 | |
| *** bana_k has joined #openstack-lbaas | 04:20 | |
| *** yuanying has joined #openstack-lbaas | 04:33 | |
| *** prabampm has joined #openstack-lbaas | 04:45 | |
| *** armax has quit IRC | 04:54 | |
| *** yamamoto has joined #openstack-lbaas | 04:56 | |
| *** bana_k has quit IRC | 05:05 | |
| *** numans has joined #openstack-lbaas | 05:27 | |
| openstackgerrit | Merged openstack/octavia: Remove oslo.serialization module requirement https://review.openstack.org/320658 | 05:59 |
| openstackgerrit | Merged openstack/octavia: Use the local /vagrant dir for copying the config and support scripts https://review.openstack.org/319871 | 05:59 |
| *** ig0r_ has joined #openstack-lbaas | 06:01 | |
| *** anilvenkata has joined #openstack-lbaas | 06:08 | |
| *** ig0r_ has quit IRC | 06:17 | |
| *** rcernin has joined #openstack-lbaas | 06:21 | |
| *** perelman has joined #openstack-lbaas | 06:30 | |
| *** Alex_Stef has joined #openstack-lbaas | 06:30 | |
| *** perelman has quit IRC | 06:38 | |
| *** perelman has joined #openstack-lbaas | 06:39 | |
| *** perelman has joined #openstack-lbaas | 06:39 | |
| *** meghal has quit IRC | 06:46 | |
| *** bana_k has joined #openstack-lbaas | 06:46 | |
| *** tesseract has joined #openstack-lbaas | 06:47 | |
| *** kevo has joined #openstack-lbaas | 06:48 | |
| *** bana_k has quit IRC | 07:16 | |
| *** kobis has joined #openstack-lbaas | 07:23 | |
| *** mixos has quit IRC | 07:25 | |
| *** jschwarz has joined #openstack-lbaas | 07:28 | |
| *** ihrachys has joined #openstack-lbaas | 07:28 | |
| *** chlong has quit IRC | 07:46 | |
| *** daneyon has quit IRC | 07:50 | |
| *** nokes has joined #openstack-lbaas | 07:53 | |
| *** eezhova has joined #openstack-lbaas | 08:03 | |
| *** kevo has quit IRC | 08:05 | |
| *** ig0r_ has joined #openstack-lbaas | 08:20 | |
| *** dmk0202 has joined #openstack-lbaas | 08:28 | |
| *** _laco has quit IRC | 08:32 | |
| *** kevo has joined #openstack-lbaas | 08:34 | |
| *** yamamoto_ has joined #openstack-lbaas | 08:35 | |
| *** yamamoto has quit IRC | 08:38 | |
| *** daneyon has joined #openstack-lbaas | 08:52 | |
| *** daneyon has quit IRC | 08:55 | |
| *** daneyon has joined #openstack-lbaas | 08:56 | |
| *** amotoki has quit IRC | 09:00 | |
| *** numans has quit IRC | 09:05 | |
| *** yuanying has quit IRC | 09:08 | |
| *** yuanying has joined #openstack-lbaas | 09:10 | |
| *** numans has joined #openstack-lbaas | 09:35 | |
| *** madhu_ak has quit IRC | 09:40 | |
| *** amotoki has joined #openstack-lbaas | 09:47 | |
| *** reedip_ has joined #openstack-lbaas | 10:15 | |
| *** reedip_ has quit IRC | 10:17 | |
| *** al_loew has joined #openstack-lbaas | 10:18 | |
| *** yamamoto_ has quit IRC | 10:28 | |
| *** eezhova has quit IRC | 10:50 | |
| *** eezhova has joined #openstack-lbaas | 10:54 | |
| openstackgerrit | Elena Ezhova proposed openstack/neutron-lbaas: Add operating and provisioning status to RESOURCE_ATTRIBUTE_MAP https://review.openstack.org/320893 | 10:56 |
| *** al_loew has quit IRC | 10:57 | |
| *** rtheis has joined #openstack-lbaas | 11:08 | |
| *** yamamoto has joined #openstack-lbaas | 11:11 | |
| *** prabampm has quit IRC | 11:12 | |
| *** yamamoto has quit IRC | 11:14 | |
| *** amotoki has quit IRC | 11:17 | |
| *** openstackgerrit has quit IRC | 11:18 | |
| *** openstackgerrit has joined #openstack-lbaas | 11:18 | |
| *** prabampm has joined #openstack-lbaas | 11:27 | |
| *** eezhova has quit IRC | 11:28 | |
| *** eezhova has joined #openstack-lbaas | 11:30 | |
| *** amotoki has joined #openstack-lbaas | 11:34 | |
| *** numans has quit IRC | 11:46 | |
| *** amotoki has quit IRC | 11:54 | |
| *** eezhova has quit IRC | 12:05 | |
| *** yamamoto has joined #openstack-lbaas | 12:13 | |
| *** numans has joined #openstack-lbaas | 12:33 | |
| *** cody-somerville_ has joined #openstack-lbaas | 12:40 | |
| *** eezhova has joined #openstack-lbaas | 12:42 | |
| *** matt-borland has joined #openstack-lbaas | 12:54 | |
| *** ducttape_ has joined #openstack-lbaas | 12:58 | |
| *** ig0r__ has joined #openstack-lbaas | 13:06 | |
| *** ig0r_ has quit IRC | 13:08 | |
| *** numans has quit IRC | 13:09 | |
| *** ducttape_ has quit IRC | 13:09 | |
| *** numans has joined #openstack-lbaas | 13:09 | |
| *** BjoernT has joined #openstack-lbaas | 13:42 | |
| *** BjoernT is now known as Bjoern_zZzZzZzZ | 13:42 | |
| *** yamamoto has quit IRC | 13:43 | |
| openstackgerrit | Elena Ezhova proposed openstack/neutron-lbaas: WIP Refactor gate hook https://review.openstack.org/320998 | 13:47 |
| openstackgerrit | Elena Ezhova proposed openstack/neutron-lbaas: WIP Add gate hook for the functional job https://review.openstack.org/320999 | 13:47 |
| *** links has quit IRC | 13:48 | |
| openstackgerrit | Miguel Angel Ajo proposed openstack/octavia: Allow deployment with pre-generated ssh keys and certificates https://review.openstack.org/321002 | 13:53 |
| openstackgerrit | Miguel Angel Ajo proposed openstack/octavia: Allow deployment with pre-generated ssh keys and certificates https://review.openstack.org/321002 | 13:57 |
| *** Bjoern_zZzZzZzZ is now known as BjoernT | 14:00 | |
| openstackgerrit | Miguel Angel Ajo proposed openstack/octavia: Use devstack RABBIT_HOST to point to rabbit, instead of localhost https://review.openstack.org/321009 | 14:01 |
| *** yamamoto has joined #openstack-lbaas | 14:04 | |
| *** ducttape_ has joined #openstack-lbaas | 14:04 | |
| *** ducttape_ has quit IRC | 14:09 | |
| *** ducttape_ has joined #openstack-lbaas | 14:09 | |
| *** mixos has joined #openstack-lbaas | 14:19 | |
| *** yamamoto has quit IRC | 14:19 | |
| *** diogogmt has quit IRC | 14:20 | |
| *** anilvenkata has quit IRC | 14:27 | |
| openstackgerrit | Elena Ezhova proposed openstack/neutron-lbaas: WIP Add gate hook for the functional job https://review.openstack.org/320999 | 14:27 |
| openstackgerrit | Elena Ezhova proposed openstack/neutron-lbaas: WIP Refactor gate hook https://review.openstack.org/320998 | 14:27 |
| *** yamamoto has joined #openstack-lbaas | 14:29 | |
| *** ihrachys_ has joined #openstack-lbaas | 14:29 | |
| *** ihrachys has quit IRC | 14:31 | |
| *** yamamoto has quit IRC | 14:32 | |
| *** yamamoto has joined #openstack-lbaas | 14:32 | |
| *** ihrachys_ has quit IRC | 14:34 | |
| *** ihrachys has joined #openstack-lbaas | 14:34 | |
| ajo_ | ihrachys, you may want to have an eye on https://review.openstack.org/#/c/318946/2 | 14:35 |
| *** jschwarz has quit IRC | 14:36 | |
| ihrachys | ajo_: I already did | 14:38 |
| ihrachys | ajo_: note I already commented on RPC issue before | 14:39 |
| ihrachys | ajo_: but that's what they do right now, so I removed my -1 since the goal of the patch is not to optimize the flow | 14:39 |
| ihrachys | long term, they should introduce their own notification driver | 14:39 |
| ihrachys | that would handle all locally | 14:40 |
| ajo_ | ihrachys, ack, btw I thought this was neutron channel somehow :) this is lbaas | 14:40 |
| *** yamamoto has quit IRC | 14:41 | |
| *** yamamoto has joined #openstack-lbaas | 14:42 | |
| ihrachys | ouch | 14:43 |
| *** ajmiller__ has quit IRC | 14:43 | |
| *** cody-somerville_ has quit IRC | 14:45 | |
| *** ajmiller has joined #openstack-lbaas | 14:45 | |
| *** yamamoto has quit IRC | 14:49 | |
| *** fnaval has quit IRC | 14:50 | |
| *** fnaval has joined #openstack-lbaas | 15:03 | |
| *** amotoki has joined #openstack-lbaas | 15:03 | |
| *** mixos has quit IRC | 15:08 | |
| *** ri0 has joined #openstack-lbaas | 15:09 | |
| *** diogogmt has joined #openstack-lbaas | 15:19 | |
| fnaval | review requested please: https://review.openstack.org/#/c/317692/ | 15:19 |
| fnaval | https://review.openstack.org/#/c/310326/ | 15:19 |
| fnaval | thanks in advance! | 15:20 |
| *** amotoki_ has joined #openstack-lbaas | 15:21 | |
| *** numans has quit IRC | 15:24 | |
| *** amotoki has quit IRC | 15:24 | |
| *** eezhova has quit IRC | 15:26 | |
| *** eezhova has joined #openstack-lbaas | 15:27 | |
| *** TrevorV has joined #openstack-lbaas | 15:29 | |
| *** cody-somerville has joined #openstack-lbaas | 15:30 | |
| *** eezhova has quit IRC | 15:30 | |
| *** mixos has joined #openstack-lbaas | 15:31 | |
| *** eezhova has joined #openstack-lbaas | 15:32 | |
| *** amotoki_ has quit IRC | 15:34 | |
| *** ri0 has quit IRC | 15:39 | |
| *** ihrachys has quit IRC | 15:41 | |
| openstackgerrit | Miguel Angel Ajo proposed openstack/octavia: Allow deployment with pre-generated ssh keys and certificates https://review.openstack.org/321002 | 15:41 |
| *** meghal has joined #openstack-lbaas | 15:42 | |
| openstackgerrit | Franklin Naval proposed openstack/neutron-lbaas: WIP: Neutron-LBaaSv2: Network Topology Tests https://review.openstack.org/319512 | 15:43 |
| *** armax has joined #openstack-lbaas | 15:43 | |
| *** kodokuu has joined #openstack-lbaas | 15:44 | |
| *** eezhova has quit IRC | 15:45 | |
| *** ri0 has joined #openstack-lbaas | 15:46 | |
| *** yamamoto has joined #openstack-lbaas | 15:48 | |
| *** meghal has quit IRC | 15:49 | |
| *** meghal has joined #openstack-lbaas | 15:49 | |
| *** dmk0202 has quit IRC | 15:49 | |
| *** madhu_ak has joined #openstack-lbaas | 15:50 | |
| openstackgerrit | Franklin Naval proposed openstack/neutron-lbaas: Neutron-LBaaSv2: TempestPlugin https://review.openstack.org/321087 | 15:51 |
| *** yamamoto has quit IRC | 15:53 | |
| *** kobis has quit IRC | 15:55 | |
| *** ri0 has quit IRC | 15:55 | |
| openstackgerrit | Franklin Naval proposed openstack/neutron-lbaas: Neutron-LBaaSv2: TempestPlugin https://review.openstack.org/321087 | 15:57 |
| *** cody-somerville has quit IRC | 16:01 | |
| *** cody-somerville has joined #openstack-lbaas | 16:01 | |
| *** cody-somerville has quit IRC | 16:02 | |
| *** cody-somerville has joined #openstack-lbaas | 16:02 | |
| *** cody-somerville has quit IRC | 16:02 | |
| *** cody-somerville has joined #openstack-lbaas | 16:02 | |
| *** woodster_ has joined #openstack-lbaas | 16:03 | |
| *** rcernin has quit IRC | 16:04 | |
| *** bana_k has joined #openstack-lbaas | 16:09 | |
| *** links has joined #openstack-lbaas | 16:10 | |
| *** tesseract has quit IRC | 16:13 | |
| *** cody-somerville has quit IRC | 16:16 | |
| *** cody-somerville has joined #openstack-lbaas | 16:21 | |
| *** cody-somerville has joined #openstack-lbaas | 16:21 | |
| *** kevo has quit IRC | 16:21 | |
| *** cody-somerville has quit IRC | 16:22 | |
| *** cody-somerville has joined #openstack-lbaas | 16:24 | |
| openstackgerrit | Sergey Belous proposed openstack/octavia: Small cleanup of plugin.sh https://review.openstack.org/321108 | 16:28 |
| *** madhu_ak has quit IRC | 16:34 | |
| *** nokes has quit IRC | 16:34 | |
| *** bana_k has quit IRC | 16:56 | |
| *** kevo has joined #openstack-lbaas | 17:00 | |
| *** ducttape_ has quit IRC | 17:08 | |
| *** ducttape_ has joined #openstack-lbaas | 17:08 | |
| *** ducttape_ has quit IRC | 17:10 | |
| *** ducttape_ has joined #openstack-lbaas | 17:10 | |
| *** bana_k has joined #openstack-lbaas | 17:23 | |
| *** eezhova has joined #openstack-lbaas | 17:30 | |
| *** BjoernT is now known as Bjoern_zZzZzZzZ | 17:36 | |
| *** piet has joined #openstack-lbaas | 17:37 | |
| *** Bjoern_zZzZzZzZ is now known as BjoernT | 17:40 | |
| *** fawadkhaliq has joined #openstack-lbaas | 17:43 | |
| *** ig0r__ has quit IRC | 17:51 | |
| *** links has quit IRC | 17:52 | |
| *** alhu has joined #openstack-lbaas | 18:03 | |
| *** _ducttape_ has joined #openstack-lbaas | 18:11 | |
| TrevorV | Hey dougwig any thoughts on the spec requirement for the single-create? | 18:13 |
| TrevorV | Kinda waiting on that to get a +A | 18:14 |
| *** ducttape_ has quit IRC | 18:15 | |
| johnsom | TrevorV point me to it | 18:21 |
| TrevorV | johnsom https://review.openstack.org/#/c/257201/ | 18:22 |
| *** fawadkhaliq has quit IRC | 18:23 | |
| johnsom | I think you are good as is. Just need the reviews. | 18:24 |
| TrevorV | johnsom yeah, just wanted an official 'yay or nay' | 18:25 |
| TrevorV | I mean, we didn't need the spec when brandon started it forever ago, but if I need one now, I need to knock that out. | 18:25 |
| TrevorV | Ya know? | 18:25 |
| johnsom | I see no need for a spec for this. We have the approved rfe which should be good enough | 18:25 |
| TrevorV | johnsom who worked on the "no_monitor" status for octavia? | 18:29 |
| TrevorV | Do you remember? | 18:29 |
| TrevorV | It looks like you added it, or at least the annotation says it was you | 18:30 |
| johnsom | TrevorV I did | 18:30 |
| TrevorV | Do I have to enable it in the octavia conf? | 18:30 |
| a2hill | I agree with johnsom, I just would hate for it to be held up last second because of an opinion when it could have been worked on sooner | 18:30 |
| a2hill | If dougwig could chime in here to give us a solid one way or the other answer i'd appreciate it | 18:31 |
| johnsom | TrevorV No, it's on now | 18:31 |
| TrevorV | Default behavior when doing a member create is to put operating_status as "OFFLINE" | 18:31 |
| TrevorV | I don't even see where it checks for the health monitor | 18:31 |
| johnsom | TrevorV That was a long time ago, let me have a look | 18:32 |
| TrevorV | Yeah, no problem | 18:32 |
| TrevorV | johnsom I'm thinking that change was overwritten at some point, probably by a shitty rebase from me when adding single-create | 18:32 |
| johnsom | TrevorV https://github.com/openstack/octavia/blob/master/octavia/controller/healthmanager/update_db.py#L173 | 18:33 |
| johnsom | So, when the haproxy reports "NO_CHECK" health manager updates the member to NO_MONITOR | 18:34 |
| *** mixos has quit IRC | 18:35 | |
| TrevorV | I see, its not added as NO_MONITOR when creating the member. | 18:35 |
| TrevorV | That complicates things on the neutron lbaas side then. | 18:35 |
| johnsom | Nope. Maybe it should | 18:35 |
| *** ajmiller has quit IRC | 18:35 | |
| johnsom | Open to dialog on that | 18:35 |
| TrevorV | Oh shit, its wednesday | 18:36 |
| TrevorV | I have an agenda item to ad | 18:36 |
| TrevorV | add*** | 18:36 |
| TrevorV | You have a link? | 18:36 |
| johnsom | https://wiki.openstack.org/wiki/Octavia/Weekly_Meeting_Agenda | 18:37 |
| *** mixos has joined #openstack-lbaas | 18:37 | |
| TrevorV | I added "Merging the Cores" as a topic, johnsom | 18:38 |
| johnsom | Didn't we talk about that two weeks ago? | 18:38 |
| *** piet has quit IRC | 18:39 | |
| TrevorV | We brought it up as a possibility, but I don't think we thought it was going to happen | 18:39 |
| TrevorV | With how little traction we've had on reviews over the past few weeks, I want to say it needs to happen sooner than later. | 18:40 |
| johnsom | wrist slapping heard.... | 18:40 |
| *** eezhova has quit IRC | 18:41 | |
| TrevorV | Its not only you, johnsom ha ha, but we've got people like a2hill who can't +2 something in Octavia, and we've got rm_work who can't +2 things in Neutron LBaaS | 18:41 |
| johnsom | Yeah, understand | 18:42 |
| TrevorV | Yeah | 18:42 |
| TrevorV | So back to the NO_MONITOR talk johnsom | 18:48 |
| TrevorV | I think I'll add a patch that does a db check for a HM | 18:48 |
| TrevorV | That way we add it as "NO_MONITOR" initially | 18:49 |
| TrevorV | The heartbeat updates will still work if someone "deletes" a HM, right | 18:49 |
| TrevorV | So that should still be great. | 18:49 |
| johnsom | Yeah, I think that would be fine. Just make sure you don't blindly set it to "NO_MONITOR" as a pool can have a HM before a member is added | 18:50 |
| *** piet_ has joined #openstack-lbaas | 18:51 | |
| TrevorV | Yeah, I was going to do a check in the "prepare" methods | 18:52 |
| TrevorV | Basically "if pool has hm, 'OFFLINE', else 'NO_MONITOR'" type of thing | 18:53 |
| johnsom | Cool | 18:53 |
| *** ajmiller has joined #openstack-lbaas | 18:57 | |
| *** ajmiller_ has joined #openstack-lbaas | 18:59 | |
| *** ajmiller has quit IRC | 19:01 | |
| *** ajmiller_ has quit IRC | 19:10 | |
| *** ajmiller has joined #openstack-lbaas | 19:16 | |
| *** _ducttape_ has quit IRC | 19:17 | |
| *** eezhova has joined #openstack-lbaas | 19:20 | |
| *** meghal1 has joined #openstack-lbaas | 19:22 | |
| *** meghal has quit IRC | 19:24 | |
| *** daneyon has quit IRC | 19:26 | |
| *** ajmiller has quit IRC | 19:27 | |
| *** ajmiller has joined #openstack-lbaas | 19:27 | |
| *** ducttape_ has joined #openstack-lbaas | 19:28 | |
| *** ajmiller_ has joined #openstack-lbaas | 19:30 | |
| *** alhu has quit IRC | 19:30 | |
| *** meghal1 has quit IRC | 19:31 | |
| *** meghal has joined #openstack-lbaas | 19:31 | |
| *** alhu has joined #openstack-lbaas | 19:31 | |
| *** ajmiller__ has joined #openstack-lbaas | 19:32 | |
| *** ajmiller_ has quit IRC | 19:33 | |
| *** daneyon has joined #openstack-lbaas | 19:33 | |
| *** ajmiller has quit IRC | 19:33 | |
| *** cody-somerville has quit IRC | 19:49 | |
| *** TrevorV has quit IRC | 19:52 | |
| johnsom | Octavia meeting starting in five minutes on #openstack-meeting-alt | 19:55 |
| *** piet_ has quit IRC | 19:56 | |
| *** mixos has quit IRC | 19:58 | |
| *** meghal1 has joined #openstack-lbaas | 19:58 | |
| *** TrevorV|Home has joined #openstack-lbaas | 19:58 | |
| *** mixos has joined #openstack-lbaas | 19:59 | |
| *** mixos has quit IRC | 19:59 | |
| *** meghal has quit IRC | 20:00 | |
| openstackgerrit | Jeffrey Longstaff proposed openstack/neutron-lbaas: Modified F5 LBaaS driver to reflect current F5 service provider implmenetation. Changed name to avoid plugin label. https://review.openstack.org/321232 | 20:44 |
| *** minwang2 has joined #openstack-lbaas | 20:52 | |
| Frito | 0/ | 21:01 |
| alhu | bana_k: what i understood is that, the transaction lock is held in db | 21:01 |
| johnsom | bana_k The spares are allocated from the list available in the database. The check for spare and allocate is inside a single transaction | 21:01 |
| bana_k | johnsom got it ! | 21:01 |
| bharathm | johnsom: Following up on bana_k question.. What if the spare check is done by two HK instances at the same time and both end up staring spare instances ? | 21:02 |
| johnsom | There have been patches where the check and allocation were separate, but I think I have fixed all of those | 21:02 |
| alhu | bharathm: one shall fail, right? | 21:02 |
| johnsom | bharathm Ah, yes, spares check for building additional spares. Not sure how tightly we have that locked, but one or two more spares should not be a big deal | 21:03 |
| alhu | or are you referring the scenario that both break off the transaction and continue processing? | 21:03 |
| johnsom | It's allocation of the spares that we really have to be careful with | 21:03 |
| alhu | how can we prevent this? | 21:03 |
| alhu | in this case, it's just a spare, but not sure any other cases | 21:04 |
| bharathm | Johnsom: got it.. Yeah I see there is a slight possiblity of starting up additional spare instances than the set spare_pool_size.. Allocation should be handled fine | 21:04 |
| johnsom | Allocation of spares we manage carefully through database transactions. | 21:04 |
| johnsom | Outside of spare creation, we are very careful. If we need to boot an amphora (zero spares pool size for example) you will note that the flow never puts it in the available state, it goes direct to allocated. | 21:05 |
| bana_k | johnsom allocation is done by the controller worker at the time of build right ? | 21:05 |
| johnsom | For spares pool, the allocation is handled inside one transaction. | 21:05 |
| johnsom | allocation is done by the controller at LB create time | 21:06 |
| bana_k | o | 21:06 |
| bana_k | ok | 21:06 |
| xgerman | yeah, having a few more spares shouldn’t be an issue | 21:06 |
| johnsom | We attempt to allocate, if none available, we build | 21:06 |
| *** TrevorV|Home has quit IRC | 21:07 | |
| bana_k | thanks all for the clarification | 21:07 |
| alhu | thank you, guys, for the answers!! | 21:07 |
| johnsom | Sure. I spent some time on that to make sure the LB create path was multi-controller safe | 21:07 |
| xgerman | and as I said we have multiple of everything running in “production" | 21:08 |
| xgerman | or at least our test systems are | 21:08 |
| alhu | xgerman: how many instances have you tested? | 21:09 |
| bharathm | alhu: I tested creating 60 LBs on a multi node cluster | 21:10 |
| xgerman | we have QA for that :-) But they tested 3 of everyhting | 21:10 |
| xgerman | aka 3*worker, h, hm, api | 21:10 |
| alhu | i c | 21:10 |
| bharathm | As xgerman said 3x of all 4 Octavia components (api, cw, hm, hk) | 21:10 |
| alhu | ipv6 should work, right? | 21:10 |
| xgerman | but we tested liberty — so with mistake YMMV | 21:11 |
| xgerman | mitaka | 21:11 |
| johnsom | alhu IPv6 is broken | 21:11 |
| alhu | :( | 21:11 |
| johnsom | I recently discovered that | 21:11 |
| johnsom | Someone, they know who they are, cut some corners in some code. It is fixable, just currently broken. | 21:12 |
| johnsom | The core load balancing element can support it, it's just the Octavia code has some bugs there | 21:12 |
| alhu | i c | 21:13 |
| alhu | johnsom: thanks | 21:14 |
| johnsom | alhu NP | 21:14 |
| johnsom | a2hill ping | 21:21 |
| *** minwang2 has quit IRC | 21:26 | |
| kevo | Trying to bring up the octavia database in a non devstack environment. The database is created in mysql. Then, in octavia/db/migration, i used octavia-db-manage upgrade head, but receive this error. https://gist.github.com/anonymous/60eb362c2aeaffade29653ea949d67ec I have the connection field in octavia.conf as connection = mysql+pymysql://root:password@controller/octavia. What am I missing? | 21:28 |
| *** matt-borland has quit IRC | 21:30 | |
| *** kodokuu has quit IRC | 21:35 | |
| *** ducttape_ has quit IRC | 21:37 | |
| *** rtheis has quit IRC | 21:37 | |
| *** cody-somerville_ has joined #openstack-lbaas | 21:42 | |
| a2hill | johnsom: pong | 21:46 |
| *** daneyon has quit IRC | 21:49 | |
| *** daneyon has joined #openstack-lbaas | 21:50 | |
| a2hill | johnsom: you pinging regarding that bug? | 21:53 |
| *** minwang2 has joined #openstack-lbaas | 21:53 | |
| johnsom | I wanted to talk about the https://bugs.launchpad.net/octavia/+bug/1582507 | 21:54 |
| openstack | Launchpad bug 1582507 in octavia "Allow host amphora information in templater" [Undecided,New] | 21:54 |
| a2hill | Ah, alright | 21:54 |
| johnsom | Since we didn't change the template, why is this needed? | 21:55 |
| a2hill | So information can be used in custom templates. its passing information that some cases my need to use | 21:55 |
| a2hill | also, this could be potentially used by active/active when that comes around, so it could be pitched as that | 21:56 |
| a2hill | the real case, is we need to build something a bit different then upstream has, and doing it in the template is the right way, it just needs some additional info | 21:56 |
| a2hill | I was thinking about adding something there, but it would be useless, but i expected this question to come up | 21:56 |
| johnsom | Ok, I think I get you. | 21:57 |
| a2hill | to get a little more detailed for ya, we are using the vrrp_ip field on the amp to set the bind ip, and also expect to utilize other info from the amphora here | 21:57 |
| a2hill | so, passing the 'host' amphora that the config is being generated for allows us to use that additional info as we need | 21:58 |
| *** piet has joined #openstack-lbaas | 21:58 | |
| a2hill | otherwise, just a way to pass 'host' type information would be needed, i figured this was the path of least resistence | 21:58 |
| johnsom | Ok | 21:59 |
| *** eezhova has quit IRC | 22:11 | |
| *** mixos has joined #openstack-lbaas | 22:26 | |
| *** diogogmt has quit IRC | 22:27 | |
| *** minwang2_ has joined #openstack-lbaas | 22:34 | |
| *** minwang2 has quit IRC | 22:37 | |
| *** minwang2_ is now known as minwang2 | 22:37 | |
| *** ducttape_ has joined #openstack-lbaas | 22:46 | |
| *** chlong has joined #openstack-lbaas | 23:05 | |
| a2hill | johnsom: i have no reason to pass all the things, Im only translating the entire object. If you feel that some of that data could potentially be misused by a deployers template ill gladly remove it from translation | 23:09 |
| johnsom | Gotcha, yeah, those two items seemed a bit unnecessary, and the ID is one part of the "security" key (such it is with keepalived) | 23:11 |
| johnsom | a2hill They just gave me that "uneasy" feeling. Other than that and the bandit alert, I think it is fine | 23:12 |
| *** harlowja has quit IRC | 23:12 | |
| a2hill | So you think i should pull those two out? There would never be a need for it? Eh, if there is we can update it. | 23:12 |
| johnsom | I am pretty sure there is never a reason the haproxy config would need those two pieces of info. | 23:13 |
| a2hill | fair enough | 23:13 |
| johnsom | Cool | 23:14 |
| *** ducttape_ has quit IRC | 23:25 | |
| *** ducttape_ has joined #openstack-lbaas | 23:28 | |
| *** BjoernT has quit IRC | 23:30 | |
| *** piet has quit IRC | 23:32 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!