| kata-irc-bot | <xwlpt> @archana.m.shinde So if we pass the device via `-v device:device`, the device info will be exist in the req.devices and req.storages ? | 00:59 |
|---|---|---|
| kata-irc-bot | <archana.m.shinde> @xwlpt Yes | 01:00 |
| kata-irc-bot | <xwlpt> @archana.m.shinde thanks. I just found the code that add it to req.storages, but when it added to req.devices? | 01:01 |
| kata-irc-bot | <archana.m.shinde> @xwlpt Take a look at this: https://github.com/kata-containers/runtime/blob/master/virtcontainers/kata_agent.go#L1035 | 01:04 |
| kata-irc-bot | <xwlpt> `// Append container devices for block devices passed with --device.` | 01:04 |
| kata-irc-bot | <xwlpt> But I pass it with `-v device:device` | 01:05 |
| kata-irc-bot | <xwlpt> I think if I use `-v device:device`, the info will only be exist at mounts info but not device info? | 01:06 |
| kata-irc-bot | <archana.m.shinde> No there will be device info as well for volumes that are block devices | 01:07 |
| kata-irc-bot | <archana.m.shinde> see this : https://github.com/kata-containers/runtime/blob/16bd9833876859b5e7fe09f42168899d2f2a7940/virtcontainers/container.go#L494 | 01:08 |
| kata-irc-bot | <xwlpt> This is used to hotplug the device to the VM, and it will also add the device to c.devices ? And then handled by `ctrDevices = k.appendDevices(ctrDevices, c)` ? | 01:11 |
| kata-irc-bot | <xwlpt> I will have a check. | 01:12 |
| kata-irc-bot | <archana.m.shinde> @xwlpt Thats right | 01:12 |
| kata-irc-bot | <xwlpt> @archana.m.shinde Thanks for kindly help. | 01:13 |
| kata-irc-bot | <archana.m.shinde> @xwlpt sure, np. | 01:13 |
| *** igordc has joined #kata-dev | 05:14 | |
| kata-irc-bot | <xwlpt> @archana.m.shinde If the device attached manually, it should not appear in the mount ns of the container, but the question is why I can’t see it at the VM’s `/dev` directly? | 05:18 |
| *** igordc has quit IRC | 05:27 | |
| *** igordc has joined #kata-dev | 06:07 | |
| *** igordc has quit IRC | 06:07 | |
| *** gmmaha has quit IRC | 07:10 | |
| *** gmmaharaj has joined #kata-dev | 07:11 | |
| *** gmmaharaj is now known as gmmaha | 07:11 | |
| kata-irc-bot | <afrosi> Kata was installed by the scripts from tests. I need to check the combination that provokes that | 07:55 |
| *** sgarzare has joined #kata-dev | 08:01 | |
| *** sameo has joined #kata-dev | 08:31 | |
| *** davidgiluk has joined #kata-dev | 08:59 | |
| *** gwhaley has joined #kata-dev | 09:03 | |
| *** jodh has joined #kata-dev | 09:17 | |
| *** sameo has quit IRC | 09:27 | |
| *** sameo has joined #kata-dev | 09:39 | |
| *** tmhoang has joined #kata-dev | 09:55 | |
| kata-irc-bot | <xwlpt> @archana.m.shinde @eric.ernst Finally I found the device via ` nsenter -t 1 -m ls /dev/` in the VM. It is very interesting of this behavior. Why this is designed ? | 10:16 |
| *** sgarzare has quit IRC | 10:32 | |
| *** sgarzare has joined #kata-dev | 10:32 | |
| *** devimc has joined #kata-dev | 13:09 | |
| *** fuentess has joined #kata-dev | 13:50 | |
| *** jugs1 is now known as jugs | 14:07 | |
| *** sameo has quit IRC | 14:14 | |
| *** sameo has joined #kata-dev | 14:50 | |
| kata-irc-bot | <xwlpt> cc @eric.ernst | 15:03 |
| *** lpetrut has joined #kata-dev | 15:50 | |
| *** sgarzare has quit IRC | 16:13 | |
| kata-irc-bot | <jose.carlos.venegas.m> @eric.ernst @james.o.hunt@graham.whaley https://github.com/kata-containers/packaging/pull/345 I think we can skip the bump for the CI for this PR | 16:26 |
| kata-irc-bot | <jose.carlos.venegas.m> are you OK with it or prefer I bump and run CI again? | 16:26 |
| kata-irc-bot | <graham.whaley> If you didn't change anything that makes the kernel physically different (so, no change to kernel version, or config file, or make incovation etc.), then I think you are probalby OK. We've had some real activity in that repo and around that version file this past week or two. | 16:29 |
| kata-irc-bot | <graham.whaley> cc @salvador.fuentes ^^ coz we also discussed this on another PR like 5 minutes ago :slightly_smiling_face: | 16:30 |
| kata-irc-bot | <jose.carlos.venegas.m> agree | 16:32 |
| kata-irc-bot | <jose.carlos.venegas.m> we can force bump only for configs and patches directory | 16:33 |
| kata-irc-bot | <jose.carlos.venegas.m> so feel free to merge please | 16:34 |
| kata-irc-bot | <salvador.fuentes> Agree merging, is just an MD | 16:35 |
| kata-irc-bot | <eric.ernst> +1 | 16:52 |
| *** lpetrut has quit IRC | 17:14 | |
| *** lpetrut has joined #kata-dev | 17:17 | |
| *** lpetrut has quit IRC | 17:32 | |
| *** jodh has quit IRC | 18:02 | |
| *** sameo has quit IRC | 18:05 | |
| *** irclogbot_0 has joined #kata-dev | 18:10 | |
| *** gwhaley has quit IRC | 18:21 | |
| *** sameo has joined #kata-dev | 18:34 | |
| *** irclogbot_0 has quit IRC | 18:36 | |
| *** sameo has quit IRC | 19:39 | |
| kata-irc-bot | <eric.ernst> ping @gmmaharaj? | 20:07 |
| kata-irc-bot | <eric.ernst> you mentioned no changes needed for/etc/containerd/config.toml? | 20:07 |
| *** igordc has joined #kata-dev | 20:10 | |
| kata-irc-bot | <eric.ernst> @fupan - any updates on ```Based on the discussion in containerd/containerd#2916, the containerd configuration will have a field for runtime configuration files (likely to beconfig_file). However, the field is not available yet, and the shimv2 will search the configuration file in the default paths, same as kata-runtime cli): /etc/kata-containers/configuration.toml, /usr/share/defaults/kata-containers/configuration.toml. We will | 20:16 |
| kata-irc-bot | update this document once the config_file field is available.``` from https://github.com/kata-containers/documentation/blob/master/how-to/containerd-kata.md#configuration | 20:16 |
| *** davidgiluk has quit IRC | 20:17 | |
| *** igordc has quit IRC | 20:35 | |
| *** eernst has joined #kata-dev | 20:50 | |
| *** eernst has quit IRC | 20:52 | |
| *** eernst has joined #kata-dev | 20:52 | |
| *** eernst has quit IRC | 20:54 | |
| *** irclogbot_0 has joined #kata-dev | 21:07 | |
| kata-irc-bot | <gmmaharaj> @eric.ernst yeah, without any config, if you pass `--runtime=io.containerd.run.kata.v2` it will look for containerd-shim-kata-v2 to handle the runtime. | 21:16 |
| *** irclogbot_0 has quit IRC | 21:28 | |
| *** igordc has joined #kata-dev | 21:53 | |
| *** devimc has quit IRC | 22:09 | |
| *** irclogbot_0 has joined #kata-dev | 22:19 | |
| kata-irc-bot | <eric.ernst> Could use feedback here: https://github.com/opencontainers/runtime-spec/issues/1002 | 22:24 |
| kata-irc-bot | <archana.m.shinde> @xwlpt Containers run in their own mount namespace, meant for isolating them. Thats why you see the behaviour. | 22:25 |
| *** irclogbot_0 has quit IRC | 22:34 | |
| *** irclogbot_0 has joined #kata-dev | 22:35 | |
| *** fuentess has quit IRC | 22:36 | |
| *** irclogbot_0 has quit IRC | 22:44 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!