| kata-irc-bot1 | <salvador.fuentes> @eric.ernst you mean adding new jobs to jenkins? | 00:28 |
|---|---|---|
| *** jugs has quit IRC | 00:56 | |
| *** jugs has joined #kata-dev | 00:59 | |
| *** sameo has quit IRC | 01:32 | |
| *** dims has quit IRC | 01:40 | |
| *** dims has joined #kata-dev | 02:11 | |
| *** dims has quit IRC | 02:25 | |
| *** dims has joined #kata-dev | 02:33 | |
| *** stackedsax has joined #kata-dev | 03:39 | |
| *** igordc has quit IRC | 04:02 | |
| *** lpetrut has joined #kata-dev | 07:03 | |
| *** eguan has quit IRC | 07:05 | |
| *** eguan has joined #kata-dev | 07:05 | |
| *** lpetrut has quit IRC | 08:07 | |
| *** lpetrut has joined #kata-dev | 08:08 | |
| *** davidgiluk has joined #kata-dev | 08:58 | |
| *** sameo has joined #kata-dev | 09:06 | |
| *** lpetrut has quit IRC | 09:14 | |
| *** gwhaley has joined #kata-dev | 09:24 | |
| kata-irc-bot1 | <graham.whaley> @samuel.ortiz @salvador.fuentes I have a feeling the ARM CI is failing right now - @pennyzct is aware, but may be offline for new year etc. | 10:05 |
| kata-irc-bot1 | <graham.whaley> @eric.ernst - the link to the CI status page is at the top of the ci repo home page now as well, to make it easier to find... | 10:05 |
| kata-irc-bot1 | <samuel.ortiz> @graham.whaley I need to have a second look, but I think it's a golint issue. | 10:05 |
| kata-irc-bot1 | <james.o.hunt> @eric.ernst, @salvador.fuentes, @samuel.ortiz, @manohar.r.castelino - any thoughts on https://github.com/kata-containers/documentation/issues/375 ? | 11:38 |
| *** dabukalam has quit IRC | 13:12 | |
| *** stackedsax has quit IRC | 13:20 | |
| kata-irc-bot1 | <samuel.ortiz> @graham.whaley `Invalid .pullapprove.yml, check pullapprove.com for details` | 13:24 |
| kata-irc-bot1 | <samuel.ortiz> @graham.whaley That's the new pay to use pullapprove, I assume? | 13:25 |
| kata-irc-bot1 | <graham.whaley> sameo: it is - it is because the pay account does not support 'github groups', and we are (were) using them for the team acks... | 13:26 |
| kata-irc-bot1 | <graham.whaley> I was going to remove that usage, as github ack checks are now covering it, but we (@thierry and I) are so close to being able to drop pullapprove completely, that I've delayed.. | 13:27 |
| kata-irc-bot1 | <samuel.ortiz> Sounds good. | 13:27 |
| kata-irc-bot1 | <thierry> sameo: they kind of screwed us a bit in terms of choice of plans, going back and forth on what "standard" actually included. | 13:31 |
| kata-irc-bot1 | <thierry> (best part being "unlimited public repos" actually meaning 5) | 13:31 |
| kata-irc-bot1 | <samuel.ortiz> @thierry 5, unlimited, what's the difference ? ;) | 13:32 |
| kata-irc-bot1 | <samuel.ortiz> @graham.whaley @thierry Ok, thanks for the details. | 13:32 |
| *** devimc has joined #kata-dev | 13:32 | |
| kata-irc-bot1 | <graham.whaley> I updated the CI status page to make it clearer it is bust, and we should ignore pullaprove right now | 13:33 |
| kata-irc-bot1 | <graham.whaley> I think over the next few hours we will see if the Zuul WIP/SoB checks are active and working for the repos, and then I will disable pullapprove at the pullapprove end, and then PR removal of the pullapprove.yaml files (yeah, all 15 of them or so...) | 13:33 |
| *** stackedsax has joined #kata-dev | 15:01 | |
| *** devimc has quit IRC | 15:38 | |
| *** devimc has joined #kata-dev | 15:39 | |
| kata-irc-bot1 | <graham.whaley> @thierry @samuel.ortiz (for ref @salvador.fuentes @james.o.hunt) I have all the branches removing .pullapprove.yaml made and in my repos ready to PR. Given we have no issues with the Zuul SoB/WIP checks overnight, then I'll PR them tomorrow. | 17:08 |
| kata-irc-bot1 | <graham.whaley> I am strongly considering disabling the pullapprove on the 5 repos we have it active on right now, as it is reporting errors, and effectively blocking PRs. @thierry - any issue with me doing that - we know we could re-activate them tomorrow if we really needed to for instance... | 17:08 |
| kata-irc-bot1 | <thierry> nope, go ahead | 17:21 |
| kata-irc-bot1 | <thierry> I'll ask for a refund due to false advertising :) | 17:21 |
| kata-irc-bot1 | <graham.whaley> why when I nuke a repo from pullapprove... do I see a 'waiting for facebook.com' in the browser status bar...????? I don't even have fb !!! ;) | 17:25 |
| kata-irc-bot1 | <graham.whaley> @thierry repos dropped from pullapprove. CI status page updated. I suspect any broken pullapprove items will not flush out from github though, so have noted that on the status page. Let's see how we look in the morning.... | 17:30 |
| *** gwhaley has quit IRC | 18:17 | |
| kata-irc-bot1 | <claire> FYI- over in the kata-marketing channel there are links to both the *Superuser Award* nomination form and the *Community Contributor Award* nomination form. Both sets of awards will be presented at the Denver Open Infrastructure Summit. We encourage you to nominate companies that are starting to use Kata for the Superuser Awards and awesome individuals who are contributing to Kata for the Community Contributor Awards. Thanks! | 18:25 |
| *** davidgiluk has quit IRC | 20:18 | |
| kata-irc-bot1 | <eric.ernst> i'm having fun with understanding what's in /var/run/kata-containers v. /var/run/vc ; also, sbs v. vm | 20:59 |
| kata-irc-bot1 | <eric.ernst> @sebastien.boeuf --^ | 20:59 |
| kata-irc-bot1 | <eric.ernst> and looking at /sys/fs/cgroups/kata, and tryinng to figure out how to clean this up. | 21:05 |
| kata-irc-bot1 | <eric.ernst> I don't see a single kata process, but these are still busy. | 21:06 |
| kata-irc-bot1 | <mike> i had that same question yesterday, didn’t get very far with it | 21:07 |
| kata-irc-bot1 | <eric.ernst> yeah, as an example; ```Control Group Tasks %CPU Memory Input/s Output/s /kata | 21:12 |
| kata-irc-bot1 | - - 103.3M - - /kata/119f4da2a705e41d0d9553eb2f8f7a49c41525d394ca6a8f741c037c6a6368d6 - - 4.9M - - | 21:12 |
| kata-irc-bot1 | /kata/1752664177f76cc01a2d4a73a8aef4cb7a1b3770f92e490a5e1f20908aa47a4d - - 5.2M - - /kata/1a344e6c1d1262cc56d671ced5bd234aedb9b35935f77e479020483cc1fd5168 | 21:12 |
| kata-irc-bot1 | - - 5.4M - - /kata/248a704062f03980ede9f40274a441129e8d0313163002af7978e31e505c088f - - 4.6M - - | 21:12 |
| kata-irc-bot1 | /kata/2df14aedd79729bc8ae6b023baaae353dd07aa144d511e98708274334545dafc - - 8.0M - - /kata/3452cec81c0654792be381ecfd0691d6eb0f827c7ae4cefdd68f811798ab8cce | 21:12 |
| kata-irc-bot1 | - - 5.3M - - /kata/35ea4bbe3ce796cecb056346aad182018656bf547a98b30d8fa31fdba5275a1e - - 5.4M - - | 21:12 |
| kata-irc-bot1 | /kata/4058233ce8fc7627177ea879b2fc9ee9296c206fe796d5787325e9249ee6c861 - - 2.3M - - /kata/4a6f6edc744093da4ca7f8f5211d02292d32ae7124befd7d722023d4d074b393 | 21:12 |
| kata-irc-bot1 | - - 5.8M - - /kata/53d375a453fec46506bebd882caa791b761105dd5084a24e1b3eda3589ec6317 ``` | 21:12 |
| devimc | rmdir /sys/fs/cgroups/kata | 21:16 |
| devimc | I think the real path chance depending of the cgroup | 21:17 |
| devimc | for example for memory /sys/fs/cgroups/memory/kata | 21:17 |
| kata-irc-bot1 | <eric.ernst> I only see /sys/fs/cgroup/systemd/kata | 21:18 |
| kata-irc-bot1 | <eric.ernst> And I cannot remove it. | 21:18 |
| kata-irc-bot1 | <eric.ernst> nor its sub directories | 21:18 |
| devimc | did you try with rmdir ? | 21:19 |
| devimc | not rm -rf | 21:19 |
| kata-irc-bot1 | <eric.ernst> ie: ```:/sys/fs/cgroup/systemd/kata$ sudo rmdir 119f4da2a705e41d0d9553eb2f8f7a49c41525d394ca6a8f741c037c6a6368d6/ rmdir: failed to remove '119f4da2a705e41d0d9553eb2f8f7a49c41525d394ca6a8f741c037c6a6368d6/': Device or resource busy``` | 21:19 |
| kata-irc-bot1 | <jose.carlos.venegas.m> egernst: probably some process still running ? | 21:19 |
| kata-irc-bot1 | <eric.ernst> systemd-cgls shows them running - but no tasks associated with it. | 21:19 |
| devimc | yep | 21:19 |
| kata-irc-bot1 | <eric.ernst> and there aren't any kata processes running. | 21:19 |
| devimc | cat 119f4da2a705e41d0d9553eb2f8f7a49c41525d394ca6a8f741c037c6a6368d6/cgroup.procs | 21:20 |
| kata-irc-bot1 | <eric.ernst> empty | 21:20 |
| devimc | uhhm | 21:20 |
| kata-irc-bot1 | <jose.carlos.venegas.m> egernst: is there nested directories ? | 21:20 |
| kata-irc-bot1 | <jose.carlos.venegas.m> probably will need to remove from child to parent ? | 21:21 |
| kata-irc-bot1 | <eric.ernst> ```/sys/fs/cgroup/systemd/kata/119f4da2a705e41d0d9553eb2f8f7a49c41525d394ca6a8f741c037c6a6368d6$ tree . ├── cgroup.clone_children ├── cgroup.procs ├── notify_on_release ├── tasks └── vcpu ├── cgroup.clone_children ├── cgroup.procs ├── notify_on_release └── tasks 1 directory, 8 files``` | 21:21 |
| kata-irc-bot1 | <jose.carlos.venegas.m> try vcpu dir first, just in case | 21:21 |
| devimc | haha yes | 21:21 |
| kata-irc-bot1 | <eric.ernst> oh this will be fun | 21:21 |
| kata-irc-bot1 | <eric.ernst> that's it. | 21:21 |
| devimc | thanks to Wei for that | 21:22 |
| kata-irc-bot1 | <eric.ernst> bash will clean the rest then :slightly_smiling_face: | 21:22 |
| devimc | haha | 21:22 |
| kata-irc-bot1 | <eric.ernst> I'll wait to give real thanks after I figure out how to abuse and recreate this mess. | 21:22 |
| kata-irc-bot1 | <eric.ernst> do we ... use v2? can we? | 21:22 |
| devimc | no | 21:23 |
| devimc | v1 | 21:23 |
| kata-irc-bot1 | <jose.carlos.venegas.m> we use v1 I am afraid, but seems that runc too | 21:23 |
| kata-irc-bot1 | <jose.carlos.venegas.m> devimc: same for runc right ? | 21:23 |
| devimc | I think so | 21:23 |
| devimc | the problem is the host kernel | 21:24 |
| devimc | if it's too old, then v1 | 21:24 |
| *** pabelanger has left #kata-dev | 21:24 | |
| kata-irc-bot1 | <eric.ernst> old being... before 2014? | 21:26 |
| kata-irc-bot1 | <eric.ernst> or when was it added? | 21:26 |
| kata-irc-bot1 | <eric.ernst> what become easier for us with v2? | 21:28 |
| devimc | v2 Linux 4.5 | 21:31 |
| devimc | I think | 21:31 |
| kata-irc-bot1 | <mike> cgroups v2 supports unpriv users which will allow unpriv k8s to run, that video rico posted in general talks about it a bit, good vid | 22:20 |
| kata-irc-bot1 | <mike> i need to brush up on my cgroup knowledge, didnt even know there was a v2 | 22:21 |
| *** devimc has quit IRC | 22:25 | |
| kata-irc-bot1 | <mike> https://lwn.net/Articles/679786/ good read | 22:52 |
| kata-irc-bot1 | <manohar.r.castelino> Pasting this issue here to get higher visibility | 23:16 |
| kata-irc-bot1 | <manohar.r.castelino> https://github.com/kata-containers/runtime/issues/1203 | 23:16 |
| kata-irc-bot1 | <manohar.r.castelino> vsock no longer works on ubuntu.. so firecracker will not work | 23:16 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!