| rm_work | then why were we doing it? O_o | 00:08 |
|---|---|---|
| rm_work | ah also, i think i see why these weren't set in the element -- they need to be run inside the netns? | 00:09 |
| rm_work | are these sysctl values unique to a netns? i didn't realize that | 00:09 |
| rm_work | i always thought it was system-wide | 00:09 |
| johnsom | Netns specific | 00:11 |
| rm_work | ok | 00:12 |
| rm_work | but ... we just don't need them in general? | 00:12 |
| johnsom | rm_work Sorry, I was done for the day a few hours ago. | 00:22 |
| rm_work | ah lol | 00:23 |
| johnsom | The netns gets a blank slate, so we have to set them there. Most of those a perf related for the tenant traffic and don't matter much for the agent. | 00:23 |
| rm_work | yeah i guess you've been working way earlier in the day | 00:23 |
| johnsom | Yeah, another 6am day | 00:23 |
| rm_work | right i mean i know they don't matter for the agent, i meant did we need them at all | 00:23 |
| rm_work | so yeah, ok, that's why they're being set at that point | 00:23 |
| rm_work | is the same true for kernel modules tho? O_o that is surprising too | 00:23 |
| rm_work | didn't think those were netns specific | 00:24 |
| johnsom | Those should not be netns specific.... What are you looking at, I'm context switching here | 00:25 |
| rm_work | plug.py | 00:25 |
| rm_work | some of the stuff we do as part of the vip plugging | 00:25 |
| rm_work | i don't think "load a kernel module" needs to be part of that <_< | 00:25 |
| rm_work | since i'm basically refactoring all of plug.py right now | 00:26 |
| rm_work | i want to remove as much of this from plug-time as possible and put it in startup or image build | 00:26 |
| rm_work | so like... WHEN we create the netns, i can have it run these sysctl commands... and when we build the image, i can have it add this ip_vs kernel module to boot | 00:27 |
| rm_work | it doesn't need to be run JIT on a vip plug | 00:27 |
| rm_work | oh rofl ok, i see... it actually creates the namespace in the plug | 00:29 |
| johnsom | Right | 00:29 |
| rm_work | ALSO not sure why | 00:29 |
| rm_work | why not just create it in the image build? | 00:29 |
| rm_work | it's a constant | 00:29 |
| johnsom | There was some reason, I think it is namespace creation time, then we also write it out for reboot scenarios | 00:29 |
| rm_work | we don't re-run plug on a reboot tho <_< | 00:30 |
| rm_work | IMO we can create the netns and set up these sysctl values during image creation | 00:30 |
| johnsom | Right, we don't which is why it also writes it out. | 00:30 |
| rm_work | yeah so .... | 00:30 |
| rm_work | why not just... | 00:31 |
| rm_work | have it already exist and be written out | 00:31 |
| johnsom | Yeah, a bit of this has been fairly "organicly" grown | 00:33 |
| rm_work | k | 00:34 |
| johnsom | I think the netns stuff was in the plug based on that old idea of more than one LB per amp | 00:35 |
| johnsom | Each would need it's own netns, etc. | 00:35 |
| johnsom | bonus points if you do not enable conntrack unless it's a UDP listener.... | 00:37 |
| johnsom | I thought it was that way, but that may have been something I just gave up on for timing reasons | 00:38 |
| johnsom | Ok, afk again | 00:39 |
| rm_work | soooo research is showing me that these ip forwarding sysctl values are NOT namespace aware anyway <_< | 00:42 |
| rm_work | BUT actual testing shows they obviously are | 00:43 |
| rm_work | so 100% of the docs i've found on the internet showing enabling them outside of the netns for use inside the netns... are wrong | 00:43 |
| rm_work | for example: https://coderwall.com/p/uf_44a/quick-ip-netns | 00:47 |
| rm_work | i can't find any documentation about where sysctl values are written out for network namespaces | 00:52 |
| johnsom | Don’t believe everything you read on the interwebs! | 01:39 |
| johnsom | There is an obscure kernel doc that talks to it. I am pretty sure the is what I referenced when dustin proposed that patch calling out they were not being set. | 01:40 |
| *** goldyfruit has quit IRC | 02:24 | |
| *** goldyfruit has joined #openstack-lbaas | 02:29 | |
| rm_work | yeah i just tested with reality | 02:42 |
| rm_work | seems more accurate | 02:42 |
| rm_work | is there really any downside to enabling ipv6 forwarding if there's no ipv6 interface? | 02:43 |
| *** goldyfruit has quit IRC | 02:53 | |
| *** threestrands has joined #openstack-lbaas | 02:54 | |
| rm_work | for now i am assuming no downside | 03:16 |
| rm_work | and moving that stuff out of the plug | 03:17 |
| rm_work | should speed up AND stabilize that call a little bit | 03:17 |
| rm_work | though what did you mean when you said: | 03:19 |
| rm_work | [16:47:43] johnsom:We should not need forwarding | 03:19 |
| rm_work | we DO need forwarding inside the netns, right? | 03:20 |
| rm_work | did you mean "not for the agent to work"? which, yeah, i am just reusing that element to do the config for the netns | 03:20 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: Allow multiple VIPs per LB https://review.opendev.org/660239 | 03:21 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: Allow multiple VIPs per LB https://review.opendev.org/660239 | 03:23 |
| *** yamamoto has quit IRC | 04:09 | |
| *** threestrands has quit IRC | 04:12 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: Allow multiple VIPs per LB https://review.opendev.org/660239 | 04:14 |
| rm_work | ^^ is very promising so far :D | 04:15 |
| rm_work | "just worked" in SINGLE topo, with+without additional VIPs | 04:15 |
| rm_work | trying in ACTIVE_STANDBY now | 04:16 |
| rm_work | works with one VIP.... | 04:18 |
| rm_work | ugh, second VIP doesn't come up via keepalived T_T | 04:23 |
| rm_work | first one still does tho... | 04:23 |
| rm_work | maybe something to do with keepalived and ipv6? | 04:23 |
| rm_work | aha! got it | 04:34 |
| rm_work | yep, ipv4+ipv6 don't coexist in a single block, but it's an easy fix | 04:34 |
| *** yamamoto has joined #openstack-lbaas | 04:43 | |
| *** yamamoto has quit IRC | 04:54 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: Allow multiple VIPs per LB https://review.opendev.org/660239 | 05:09 |
| rm_work | cooooool | 05:10 |
| rm_work | ok, now just centos I think :D | 05:10 |
| rm_work | gthiemonge: let me know when you have a sec to sync up ^_^ | 05:10 |
| *** yamamoto has joined #openstack-lbaas | 05:13 | |
| *** luksky has joined #openstack-lbaas | 05:44 | |
| *** ccamposr has joined #openstack-lbaas | 06:03 | |
| *** ccamposr has quit IRC | 06:05 | |
| *** ccamposr has joined #openstack-lbaas | 06:05 | |
| *** sapd1_x has joined #openstack-lbaas | 06:21 | |
| *** pcaruana has joined #openstack-lbaas | 06:35 | |
| *** tesseract has joined #openstack-lbaas | 07:09 | |
| *** Emine has joined #openstack-lbaas | 07:52 | |
| *** rcernin has quit IRC | 07:55 | |
| *** bcafarel has quit IRC | 08:13 | |
| *** sapd1_x has quit IRC | 08:15 | |
| *** bcafarel has joined #openstack-lbaas | 08:26 | |
| *** trident has quit IRC | 08:35 | |
| *** trident has joined #openstack-lbaas | 08:37 | |
| *** emine__ has joined #openstack-lbaas | 08:51 | |
| *** Emine has quit IRC | 08:52 | |
| *** sapd1_x has joined #openstack-lbaas | 09:23 | |
| *** gcheresh has joined #openstack-lbaas | 09:57 | |
| *** yamamoto has quit IRC | 10:02 | |
| *** luksky has quit IRC | 10:11 | |
| *** sapd1_x has quit IRC | 10:18 | |
| *** gcheresh has quit IRC | 10:24 | |
| *** yamamoto has joined #openstack-lbaas | 10:29 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: Allow multiple VIPs per LB https://review.opendev.org/660239 | 10:33 |
| *** yamamoto has quit IRC | 10:35 | |
| *** luksky has joined #openstack-lbaas | 10:49 | |
| *** ataraday_ has joined #openstack-lbaas | 10:55 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Add 'additional_vips' field to driver datamodel https://review.opendev.org/660238 | 11:07 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Allow multiple VIPs per LB https://review.opendev.org/660239 | 11:14 |
| rm_work | ^^ ready for basic review! :) will be working on the client and some tempest tests later today | 11:16 |
| rm_work | and gthiemonge will probably have some fixes for centos edge cases, but it works for the basic testing I've done (SINGLE/ACTIVE_STANDBY with single VIP and double VIP [primary ipv4, additional ipv6] on both Ubuntu and CentOS) | 11:17 |
| rm_work | I dread the tempest testing, as I'll literally just have to spin a ton of new LBs ... may need a whole new tempest test run <_< | 11:18 |
| rm_work | thanks for the review cgoncalves :D | 11:32 |
| cgoncalves | thanks for the release note and story ;) | 11:35 |
| *** yamamoto has joined #openstack-lbaas | 11:46 | |
| *** yamamoto has quit IRC | 11:54 | |
| *** yamamoto has joined #openstack-lbaas | 11:56 | |
| *** yamamoto has quit IRC | 12:18 | |
| *** yamamoto has joined #openstack-lbaas | 12:19 | |
| *** gthiemon1e has joined #openstack-lbaas | 12:20 | |
| *** happyhemant has joined #openstack-lbaas | 12:20 | |
| *** gthiemonge has quit IRC | 12:21 | |
| *** ianychoi_ has joined #openstack-lbaas | 12:23 | |
| *** ianychoi has quit IRC | 12:27 | |
| *** gthiemon1e has quit IRC | 12:32 | |
| *** rtjure has joined #openstack-lbaas | 12:40 | |
| *** gcheresh has joined #openstack-lbaas | 12:42 | |
| *** boden has joined #openstack-lbaas | 12:59 | |
| *** gthiemonge has joined #openstack-lbaas | 13:08 | |
| *** yamamoto has quit IRC | 13:15 | |
| *** gcheresh has quit IRC | 13:15 | |
| *** rtjure has quit IRC | 13:32 | |
| *** gthiemonge has quit IRC | 13:40 | |
| *** rtjure has joined #openstack-lbaas | 13:42 | |
| *** henriqueof has quit IRC | 13:46 | |
| *** yamamoto has joined #openstack-lbaas | 13:48 | |
| *** yamamoto has quit IRC | 13:49 | |
| *** yamamoto has joined #openstack-lbaas | 13:50 | |
| *** ricolin has joined #openstack-lbaas | 13:55 | |
| *** ccamposr has quit IRC | 13:58 | |
| *** goldyfruit has joined #openstack-lbaas | 14:28 | |
| *** sapd1_x has joined #openstack-lbaas | 15:07 | |
| *** Vorrtex has joined #openstack-lbaas | 15:34 | |
| openstackgerrit | Merged openstack/octavia master: Limit cryptsetup key RAM usage https://review.opendev.org/663784 | 15:35 |
| *** gthiemonge has joined #openstack-lbaas | 15:36 | |
| *** goldyfruit has quit IRC | 15:45 | |
| *** gthiemonge has quit IRC | 15:53 | |
| *** ramishra has quit IRC | 15:56 | |
| *** luksky has quit IRC | 16:11 | |
| *** emine__ has quit IRC | 16:19 | |
| *** yamamoto has quit IRC | 16:23 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia master: Add RHEL 8 amphora support https://review.opendev.org/638581 | 16:35 |
| *** tesseract has quit IRC | 16:40 | |
| *** boden has quit IRC | 16:46 | |
| *** yamamoto has joined #openstack-lbaas | 16:54 | |
| *** yamamoto has quit IRC | 16:59 | |
| *** boden has joined #openstack-lbaas | 17:04 | |
| *** KeithMnemonic has quit IRC | 17:09 | |
| *** sapd1_x has quit IRC | 17:19 | |
| *** ricolin has quit IRC | 17:24 | |
| *** goldyfruit has joined #openstack-lbaas | 17:48 | |
| *** luksky has joined #openstack-lbaas | 17:56 | |
| ataraday_ | johnsom, Hi! I left a comment on https://review.opendev.org/#/c/660236/ - could you check if it is valid or not? | 18:33 |
| johnsom | ataraday_ Thank you. Likely I have a bug. | 18:42 |
| *** ataraday_ has quit IRC | 19:05 | |
| *** boden has quit IRC | 19:22 | |
| *** emine__ has joined #openstack-lbaas | 19:28 | |
| *** boden has joined #openstack-lbaas | 19:33 | |
| *** boden has quit IRC | 19:33 | |
| *** emine has joined #openstack-lbaas | 19:56 | |
| *** emine__ has quit IRC | 19:59 | |
| *** goldyfruit has quit IRC | 20:02 | |
| *** emine has quit IRC | 20:04 | |
| *** emine has joined #openstack-lbaas | 20:05 | |
| *** emine__ has joined #openstack-lbaas | 20:11 | |
| *** emine has quit IRC | 20:13 | |
| *** emine has joined #openstack-lbaas | 20:15 | |
| *** emine__ has quit IRC | 20:16 | |
| *** emine has quit IRC | 20:53 | |
| *** Vorrtex has quit IRC | 20:57 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia stable/stein: Limit cryptsetup key RAM usage https://review.opendev.org/664068 | 21:46 |
| *** pcaruana has quit IRC | 21:51 | |
| *** luksky has quit IRC | 21:55 | |
| rm_work | johnsom: https://review.opendev.org/#/c/660238/ ? :D | 22:26 |
| *** goldyfruit has joined #openstack-lbaas | 22:28 | |
| openstackgerrit | Merged openstack/octavia-lib master: Add 'additional_vips' field to driver datamodel https://review.opendev.org/660238 | 22:39 |
| rm_work | wooo | 22:55 |
| rm_work | hmmm, do we have to cut a release for it before it'll work I guess? | 23:11 |
| rm_work | yeah, let's see | 23:11 |
| johnsom | yep | 23:11 |
| rm_work | doesn't look like anything is waiting | 23:11 |
| rm_work | so | 23:11 |
| rm_work | we can do that now | 23:11 |
| rm_work | ah and I had a question for you -- https://github.com/acassen/keepalived/issues/497#issuecomment-273122561 | 23:12 |
| rm_work | I did the "virtual_ipaddress_excluded" thing for the ipv6 addresses in keepalived and it seems to work in devstack | 23:12 |
| rm_work | the address comes up and i can hit it | 23:12 |
| rm_work | but i don't know if there are other side-effects to that | 23:12 |
| rm_work | the other way is a bit more complex but also doable... | 23:12 |
| johnsom | rm_work I think the virutal_ipaddress_excluded is a fine path. Have you tested if the address in this block is ifdown it fails over correctly? | 23:19 |
| rm_work | hmmm i need to do some failover tests | 23:19 |
| rm_work | i did not yet | 23:19 |
| johnsom | I think it should, but might be work a quick test. | 23:19 |
| rm_work | yeah sec | 23:19 |
| johnsom | Yeah, that is a super odd edge case, but if it passes that I think it's fine | 23:19 |
| rm_work | i'm looking for the release doc | 23:20 |
| rm_work | I am on https://docs.openstack.org/project-team-guide/release-management.html#how-to-release | 23:20 |
| johnsom | nope | 23:20 |
| johnsom | https://releases.openstack.org/reference/using.html#using-new-release-command | 23:20 |
| rm_work | but it doesn't ... actually tell me HOW in a technical sense | 23:20 |
| rm_work | ah k | 23:20 |
| rm_work | i couldn't find the link you sent earlier | 23:20 |
| johnsom | BTW, did you tell me you dropped Carlos from the release liaison and added me? I thought he was still on the release list but they didn't accept his patch. | 23:21 |
| rm_work | err | 23:21 |
| rm_work | no | 23:21 |
| rm_work | Carlos should still be it | 23:21 |
| rm_work | but now i see you there | 23:22 |
| rm_work | wat | 23:22 |
| rm_work | hold on | 23:22 |
| johnsom | https://wiki.openstack.org/w/index.php?title=CrossProjectLiaisons&diff=prev&oldid=169606 | 23:22 |
| rm_work | hmmmmmm | 23:22 |
| johnsom | You can probably add us both | 23:22 |
| rm_work | i switched you for something else | 23:22 |
| rm_work | i thought | 23:22 |
| rm_work | unless i'm misremembering, but obviously it shows me doing that so | 23:22 |
| rm_work | maybe i had the wrong field | 23:23 |
| rm_work | fixed it | 23:25 |
| johnsom | Ok | 23:26 |
| rm_work | so we're in "train" | 23:26 |
| rm_work | so that's what the release is for? | 23:26 |
| johnsom | yes | 23:27 |
| rm_work | it's going to be `train octavia-lib feature`? | 23:27 |
| johnsom | train octavia-lib feature | 23:27 |
| rm_work | or are we at a milestone | 23:27 |
| rm_work | approximately | 23:27 |
| johnsom | This week was MS1, but I would do feature so the version number is right. | 23:28 |
| rm_work | k | 23:28 |
| johnsom | in semver those are <major>.<feature>.<bugfix> | 23:28 |
| johnsom | in which number gets bumped | 23:28 |
| *** goldyfruit has quit IRC | 23:29 | |
| rm_work | ok | 23:30 |
| rm_work | https://review.opendev.org/#/c/664111/ | 23:30 |
| rm_work | don't know if that actually requires another signoff since I made it | 23:30 |
| rm_work | probably not? | 23:30 |
| rm_work | johnsom: yep, failover works (again, at least in my local devstack) | 23:35 |
| rm_work | for both addresses | 23:35 |
| rm_work | not sure if that changes if there's more layers of switches/etc in the picture | 23:35 |
| johnsom | Nope, PTL can release with no additional approvals | 23:35 |
| rm_work | or i'm off a single host | 23:35 |
| johnsom | That is fine. If the eth1:2 going down triggers the failover, I'm good with it | 23:35 |
| johnsom | No need for the multi-instance and sync stuff | 23:36 |
| johnsom | Plus it saves some bits on the wire | 23:36 |
| johnsom | Log offloading, powered by Kygo | 23:44 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!