| *** crobinso has quit IRC | 00:28 | |
| *** ericernst has quit IRC | 00:31 | |
| *** amorenoz has quit IRC | 01:13 | |
| *** amorenoz has joined #kata-dev | 01:33 | |
| *** bpradipt has joined #kata-dev | 03:55 | |
| *** bpradipt has quit IRC | 04:42 | |
| *** bpradipt has joined #kata-dev | 05:02 | |
| *** sameo has joined #kata-dev | 05:18 | |
| *** dklyle has quit IRC | 06:13 | |
| *** jodh has joined #kata-dev | 07:14 | |
| *** hashar has joined #kata-dev | 07:24 | |
| *** jodh has quit IRC | 07:35 | |
| *** jodh has joined #kata-dev | 07:36 | |
| *** davidgiluk has joined #kata-dev | 08:04 | |
| kata-irc-bot | <fidencio> @bergwolf, about the shimv2 log levels, is there a specification that we follow for that? or at least for the supported command line options? | 08:16 |
|---|---|---|
| kata-irc-bot | <bergwolf> I don’t think there is any log level specification. what are you trying to solve? | 08:34 |
| kata-irc-bot | <fidencio> The pain of shimv2 logs becoming useless, as either I have too many of those, or don't have logs at all. | 08:36 |
| kata-irc-bot | <fidencio> What I'd like to see is info logs, for instance, without having to enable the debug log level | 08:37 |
| kata-irc-bot | <fidencio> Right now, we explicitly pass "-debug" to the binary. What I'd prefer is having something as "--log-level [info|debug|warn|...]" | 08:38 |
| kata-irc-bot | <fidencio> and, preferably, have that info de-coupled from crio/containerd | 08:39 |
| kata-irc-bot | <fidencio> Actually, let me file an issue for that | 08:49 |
| kata-irc-bot | <bergwolf> Does containerd log level affect kata? Looks like we need a way for containerd to pass the log level information down to kata | 08:49 |
| kata-irc-bot | <fidencio> First things we need is shimv2 being able to accept any log level, and not only "debug" | 08:50 |
| kata-irc-bot | <bergwolf> for `--debug` we can set to `debug` level and non `--debug` we set to `info` level by default? | 08:50 |
| kata-irc-bot | <bergwolf> maybe we can add a shimv2 config option to set loglevel | 08:51 |
| kata-irc-bot | <fidencio> currently CRI-O (not sure about containerd) will pass down "-debug" to shimv2 | 08:51 |
| kata-irc-bot | <bergwolf> then shimv2 can find out the log level via config options not command line options | 08:52 |
| kata-irc-bot | <fidencio> which config option? | 08:53 |
| kata-irc-bot | <fidencio> Let me open an issue, we keep the discussion there. :slightly_smiling_face: | 08:54 |
| kata-irc-bot | <bergwolf> shimv2 API CreateTaskRequest has a `Options` field | 08:56 |
| kata-irc-bot | <bergwolf> we use it to pass kata shimv2 options | 08:56 |
| kata-irc-bot | <fidencio> Aha, that could be an option, indeed! | 08:58 |
| kata-irc-bot | <fidencio> https://github.com/kata-containers/kata-containers/issues/366 | 09:05 |
| *** mugsie has quit IRC | 10:07 | |
| *** mugsie has joined #kata-dev | 10:10 | |
| *** hashar has quit IRC | 10:47 | |
| *** hashar has joined #kata-dev | 12:22 | |
| *** vgoyal has joined #kata-dev | 12:25 | |
| *** devimc has joined #kata-dev | 12:27 | |
| *** crobinso has joined #kata-dev | 13:26 | |
| *** devimc has quit IRC | 13:29 | |
| *** devimc has joined #kata-dev | 13:29 | |
| devimc | fidencio: oi, boa tarde | 14:16 |
| devimc | fidencio: i'm trying to build a static qemu using an ubuntu 20 container on RHEL 8, but it's failing resolving *.ubuntu.com | 14:17 |
| devimc | http://jenkins.katacontainers.io/job/kata-containers-runtime-rhel-8-q35-PR/45/console | 14:17 |
| devimc | any thoughts? | 14:17 |
| devimc | Step 7/19 : RUN apt-get update && apt-get upgrade -y | 14:17 |
| devimc | ---> Running in d28a0e2078ed | 14:17 |
| devimc | Err:1 http://security.ubuntu.com/ubuntu focal-security InRelease | 14:17 |
| devimc | Temporary failure resolving 'security.ubuntu.com' | 14:17 |
| kata-irc-bot | <fidencio> Are you using a *local* el8 box or azure powered one? | 14:19 |
| kata-irc-bot | <fidencio> seems like DNS / network not working as expected | 14:20 |
| kata-irc-bot | <fidencio> Considering I can access security.ubuntu.com here without any issue | 14:20 |
| fidencio | devimc: if you can allow me to jump in that machine, I can try to debug this issue | 14:25 |
| devimc | fidencio: it's failing on azure, let me setup a VM | 14:27 |
| kata-irc-bot | <bergwolf> @fidencio could you take another look ? https://github.com/kata-containers/kata-containers/pull/362 | 14:40 |
| fidencio | @bergwolf, acked! | 14:42 |
| *** dklyle has joined #kata-dev | 14:43 | |
| *** jodh has quit IRC | 15:00 | |
| *** bpradipt has quit IRC | 15:11 | |
| *** bpradipt has joined #kata-dev | 15:14 | |
| *** hashar is now known as hasharAway | 15:24 | |
| *** devimc has quit IRC | 15:51 | |
| *** devimc has joined #kata-dev | 15:53 | |
| *** hasharAway is now known as hashar | 15:58 | |
| *** Yarboa has quit IRC | 16:14 | |
| *** Yarboa has joined #kata-dev | 16:29 | |
| *** hashar is now known as hasharAway | 17:40 | |
| *** hasharAway is now known as hashar | 19:01 | |
| *** davidgiluk has quit IRC | 19:26 | |
| *** bpradipt has quit IRC | 19:33 | |
| fidencio | devimc: is 2.0-dev supposed to support the go agent? | 19:47 |
| fidencio | seems that not, but somehow it comes as a surprise to me | 19:47 |
| kata-irc-bot | <fidencio> @bergwolf ^^ | 19:48 |
| devimc | fidencio: yeah it works, but we won't supported | 19:49 |
| devimc | *support it | 19:50 |
| fidencio | devimc: does it actually work? I mean, was it ever tested before? | 19:50 |
| devimc | rust agent will be the new agent | 19:50 |
| devimc | fidencio: uhmm I haven't tested the latest changes | 19:51 |
| fidencio | devimc: https://github.com/kata-containers/kata-containers/issues/336#issuecomment-652003721 | 19:52 |
| fidencio | take a look at case 3, the error sounds like there's no communication with the agent | 19:52 |
| devimc | fidencio: right | 19:53 |
| devimc | fidencio: did you build the image? | 19:53 |
| devimc | fidencio: could you run a quick test using the image from here http://jenkins.katacontainers.io/job/image-nightly-x86_64/lastSuccessfulBuild/artifact/artifacts/ | 19:54 |
| devimc | ? | 19:54 |
| fidencio | devimc: I can, but I'm using Fedora kernel and my own initrd which just works for the cases 1 and 2 | 19:54 |
| *** hashar is now known as hasharAway | 19:55 | |
| devimc | uhmm | 19:55 |
| fidencio | devimc: so, I'm not exactly sure how the image would make a difference | 19:55 |
| devimc | fidencio: kernel http://jenkins.katacontainers.io/job/kernel-nightly-x86_64/ | 19:55 |
| fidencio | devimc: still, why could it cause different results? | 19:56 |
| fidencio | what I'm trying to do is isolate as much variables as possibles right now | 19:56 |
| fidencio | using the very same kernel / initrd / cri-o / crictl ... meaning that the only thing that changed was the shimv2 binary | 19:57 |
| devimc | right | 19:58 |
| fidencio | btw the question is sincere and not pushing back on your suggestion, do you think the image / kernel could have some effect on this? why? | 19:59 |
| devimc | nop, I think you're right | 20:00 |
| devimc | will be the same | 20:00 |
| fidencio | okay, just to be sure I'm not going crazy here. | 20:01 |
| fidencio | I'll stop for Today and will get back to 2.0-dev branch, bisecting it | 20:01 |
| fidencio | A problem I foresee is that not every commit in that branch may be buildable | 20:02 |
| fidencio | but let's hope for the best :-) | 20:02 |
| *** Yarboa has quit IRC | 20:12 | |
| *** Yarboa has joined #kata-dev | 20:25 | |
| *** hasharAway has quit IRC | 20:28 | |
| *** hashar has joined #kata-dev | 20:29 | |
| *** devimc has quit IRC | 20:55 | |
| *** sameo has quit IRC | 21:08 | |
| *** jbryce has quit IRC | 21:15 | |
| *** jbryce has joined #kata-dev | 21:15 | |
| *** sameo has joined #kata-dev | 21:30 | |
| *** david-lyle has joined #kata-dev | 21:34 | |
| *** dklyle has quit IRC | 21:35 | |
| *** david-lyle is now known as dklyle | 21:37 | |
| *** crobinso has quit IRC | 21:39 | |
| *** hashar has quit IRC | 22:07 | |
| *** sameo has quit IRC | 23:42 | |
| *** vgoyal has quit IRC | 23:43 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!