| *** sameo has joined #kata-dev | 00:02 | |
| *** eernst has joined #kata-dev | 01:17 | |
| *** sameo has quit IRC | 01:39 | |
| *** eernst has quit IRC | 02:02 | |
| *** Bhujay has joined #kata-dev | 05:43 | |
| *** Bhujay has quit IRC | 05:44 | |
| *** Bhujay has joined #kata-dev | 05:45 | |
| *** Bhujay has quit IRC | 05:46 | |
| *** Bhujay has joined #kata-dev | 05:46 | |
| *** Bhujay has quit IRC | 07:01 | |
| *** Bhujay has joined #kata-dev | 07:10 | |
| *** Bhujay has quit IRC | 07:11 | |
| *** Bhujay has joined #kata-dev | 07:11 | |
| *** jugs has quit IRC | 08:17 | |
| *** jugs has joined #kata-dev | 08:17 | |
| *** jodh has joined #kata-dev | 08:27 | |
| *** sameo has joined #kata-dev | 08:45 | |
| kata-irc-bot | <thierry> jcvenegas: ok! I suspect it's fine to consider those machine-generated commits and exclude them from the metrics ? | 08:48 |
|---|---|---|
| kata-irc-bot | <thierry> jcvenegas: (I'd love a presentation of that automation when we can find the time. I'll trade you with a presentation of the OpenStack git-driven release automation!) | 08:50 |
| *** davidgiluk has joined #kata-dev | 09:00 | |
| *** gwhaley has joined #kata-dev | 09:30 | |
| kata-irc-bot | <xwlpt> Hi I met an issue that VM can’t be lauched. I want to get the qemu log to check why, but it seems there is no qemu logs. So how to get the qemu log ? | 09:32 |
| stefanha | xwlpt: A trick I often use is to point kata's configuration.toml at a shell script that adds extra options, strace, logging, etc before launching the real program. | 09:37 |
| stefanha | xwlpt: For example, I might add a serial console on the QEMU command-line so that it's easy to see the kernel messages from the host. | 09:37 |
| stefanha | xwlpt: For example, configuration.toml [hypervisor.qemu] path = "/path/to/qemu/kata-qemu-wrapper" | 09:41 |
| stefanha | xwlpt: And then make kata-qemu-wrapper a shell script that does: exec /home/stefanha/qemu/x86_64-softmmu/qemu-system-x86_64 "$@" -serial file:/tmp/serial.log | 09:41 |
| stefanha | xwlpt: You can add console=ttyS0 in configuration.toml's kernel_params= | 09:42 |
| stefanha | xwlpt: Now you should get boot messages in /tmp/serial.log on the host. | 09:42 |
| stefanha | xwlpt: This is useful if early boot or kata-proxy is broken. | 09:43 |
| gwhaley | xwlpt, stefanha: yep, re-directing the qemu launch via a script is a trick we used to use (eh jodh?). Also, you can enable all the 'debug=true' lines in the config toml file, and then use the system journalctl to look at the kata logs, which may also give you clues | 09:53 |
| kata-irc-bot | <james.o.hunt> @xwlpt - the most likely issue is insufficient memory to run qemu. How much memory does the system have? If you enable full debug, you should get an explanatory error from docker and in the runtime logs: https://github.com/kata-containers/documentation/blob/master/Developer-Guide.md#enable-full-debug, "journalctl -q -o cat -a -t kata-runtime > runtime.log" | 09:59 |
| kata-irc-bot | <james.o.hunt> @graham.whaley - yep, we've used a few tricks like that in the past ;) | 09:59 |
| kata-irc-bot | <xwlpt> @james.o.hunt My memory is 320G | 10:00 |
| kata-irc-bot | <james.o.hunt> @xwlpt - see what full debug gives you then. | 10:00 |
| kata-irc-bot | <xwlpt> @james.o.hunt Will have try, thanks | 10:01 |
| *** lpetrut has joined #kata-dev | 10:39 | |
| *** gwhaley has quit IRC | 11:52 | |
| *** jugs has quit IRC | 12:13 | |
| *** jugs has joined #kata-dev | 12:14 | |
| *** gwhaley has joined #kata-dev | 13:29 | |
| kata-irc-bot | <sebastien.boeuf> @fupan around? | 14:31 |
| kata-irc-bot | <sebastien.boeuf> @fupan I've just opened 2 issues related to our containerd-shim-v2 implementation: https://github.com/kata-containers/runtime/issues/1130 and https://github.com/kata-containers/runtime/issues/1131, let me know if you think one of them does not make sense | 14:47 |
| *** Bhujay has quit IRC | 15:02 | |
| kata-irc-bot | <thierry> @graham.whaley: re https://review.openstack.org/#/c/630998/ looks like your YAML was missing a couple of colons, I pushed a new patchset that should fix it | 15:38 |
| *** pabelanger has joined #kata-dev | 15:40 | |
| kata-irc-bot | <graham.whaley> ttx: ah, thanks. I wondered what you changed (was hard to tell on gerrit for me). Yeah, I guess my yaml syntax highlighter is not on where I did the edit, and I can't really test that file locally... thx! | 15:42 |
| kata-irc-bot | <graham.whaley> @thierry ^^ | 15:42 |
| kata-irc-bot | <thierry> To see what I changed: https://review.openstack.org/#/c/630998/1..2 | 15:43 |
| kata-irc-bot | <thierry> or directly https://review.openstack.org/#/c/630998/1..2/zuul.d/projects.yaml | 15:43 |
| kata-irc-bot | <thierry> I /think/ that should do it | 15:43 |
| kata-irc-bot | <graham.whaley> ah, found the way to 'diff patchsets' on the UI now ;) I've managed to avoid gerrit for a decade - guess now I have to learn to prod it a bit... | 15:44 |
| *** jodh has quit IRC | 15:58 | |
| *** eernst has joined #kata-dev | 16:31 | |
| kata-irc-bot | <graham.whaley> @samuel.ortiz wrt, does github ack counting tell you who can ack (who is missing from acking) your PRs ... it looks like it does not afaict. This is the test PR I have open: https://github.com/kata-containers/proxy/pull/138 there you can see github saying '2 acks required' and also 'codeowners required' - but, it does not have a link or list of who that might be. I suspect this is because, unless you are in the required org, you | 16:34 |
| kata-irc-bot | generally cannot see members of a 'team' in github. Thus, it might be something they'd generally advertise. So, no, not quite as slick a UI as pullapprove. I think we will be OK, as a general 'ping' will catch many folks, and then some of those can nudge the specific relevant github team for that repo. Let's see if 'actions' provides any other options. I think ack counting will be harder for zuul. The other alternative, that would require a | 16:34 |
| kata-irc-bot | maintainer and some 24/7 hardware, was https://github.com/jusx/mergeable, which is pullapprove-like. | 16:34 |
| *** dklyle has joined #kata-dev | 16:35 | |
| kata-irc-bot | <samuel.ortiz> @graham.whaley Yes, the 24/7 hardware option is the worst one for sure. | 16:37 |
| *** sameo has quit IRC | 17:06 | |
| *** stackedsax has joined #kata-dev | 17:23 | |
| *** gwhaley has quit IRC | 18:08 | |
| *** sameo has joined #kata-dev | 19:11 | |
| kata-irc-bot | <mnaser> @salvador.fuentes is it possible you changed the settings when azure was added? was it recently changed? | 19:29 |
| kata-irc-bot | <salvador.fuentes> @mnaser, I dont think so, havent touched the vexxhost config, azure was added ~2 weeks ago | 19:30 |
| kata-irc-bot | <salvador.fuentes> @mnaser, I tried testing the connection to vexxhost on the cloud configuration page and it succeeds | 19:32 |
| kata-irc-bot | <mnaser> hmm | 19:33 |
| kata-irc-bot | <salvador.fuentes> what I was wondering is if the kata-standard-4 flavor is still on your vexxhost cloud? | 19:35 |
| kata-irc-bot | <mnaser> ah, this was something that was caused because of how kata is a special case in the cloud, hopefully when we get around moving to the standardized flavors it would eliminate it all | 19:35 |
| kata-irc-bot | <mnaser> seeing them in launching state | 19:36 |
| kata-irc-bot | <mnaser> and i see them going online | 19:37 |
| kata-irc-bot | <salvador.fuentes> Yes, now I see them | 19:37 |
| *** lpetrut has quit IRC | 19:38 | |
| kata-irc-bot | <mnaser> we'll be able to give more resources if/when we decide to move towards using standardized flavors | 19:38 |
| kata-irc-bot | <mnaser> the biggest issue at the time is the /dev/sdb device which is assumed t oexist | 19:38 |
| kata-irc-bot | <mnaser> if we can kill that assumption, we can go to our normal flavors | 19:38 |
| kata-irc-bot | <salvador.fuentes> @mnaser, ok, we have that to test the devicemapper stuff, but I remember @julio.montes had a way to do it without the sdb device, will let you know if we are able to continue without that and move to standard flavors | 19:40 |
| kata-irc-bot | <salvador.fuentes> thanks | 19:40 |
| kata-irc-bot | <mnaser> great, i think that's actually how we're testing things in zuul afaik | 19:40 |
| *** davidgiluk has quit IRC | 20:08 | |
| *** sameo has quit IRC | 20:20 | |
| *** jugs has quit IRC | 20:48 | |
| *** jugs has joined #kata-dev | 20:49 | |
| *** lpetrut has joined #kata-dev | 22:34 | |
| *** lpetrut has quit IRC | 22:38 | |
| *** jugs has quit IRC | 22:51 | |
| *** jugs has joined #kata-dev | 22:53 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!