*** sameo has joined #kata-general | 02:04 | |
*** sameo has quit IRC | 02:45 | |
*** igordc has quit IRC | 04:34 | |
*** sameo has joined #kata-general | 06:59 | |
*** sameo has quit IRC | 08:02 | |
*** sgarzare has joined #kata-general | 08:28 | |
*** gwhaley has joined #kata-general | 08:53 | |
*** lpetrut has joined #kata-general | 10:34 | |
*** lpetrut has quit IRC | 11:53 | |
*** gwhaley has quit IRC | 11:59 | |
*** gwhaley has joined #kata-general | 13:18 | |
*** lpetrut has joined #kata-general | 13:40 | |
*** igordc has joined #kata-general | 14:57 | |
*** sameo has joined #kata-general | 15:15 | |
*** igordc has quit IRC | 15:17 | |
*** igordc has joined #kata-general | 15:25 | |
*** igordc has quit IRC | 15:30 | |
*** sameo has quit IRC | 15:36 | |
*** lpetrut has quit IRC | 16:38 | |
*** sgarzare has quit IRC | 17:20 | |
*** igordc has joined #kata-general | 18:15 | |
*** kata-irc-bot3 has quit IRC | 19:30 | |
*** kata-irc-bot has joined #kata-general | 19:30 | |
*** gwhaley has quit IRC | 20:25 | |
zer0def | @archana.m.shinde you happen to be around? i'm subscribing kata-containers/documentation#565 and i've brought up the `--net=none` bug in #podman some time ago, but it looks like it's conmon not returning a proper container PID from it's execution (returns -1 instead), which seems to default to creating a netns, something a user probably doesn't want with `--net=none` | 21:20 |
---|---|---|
zer0def | tl;dr: might not be exclusively a podman bug, but there's also a case of weird default behavior | 21:21 |
kata-irc-bot | <archana.m.shinde> yes, If I recall you were seeing the permission error in a setupNetNS function | 21:24 |
kata-irc-bot | <archana.m.shinde> which shouldnt be called in the first place | 21:24 |
kata-irc-bot | <archana.m.shinde> so yeah, something else leading to the final failure | 21:24 |
kata-irc-bot | <archana.m.shinde> would make sense to file an issue for it in podman and then go from there | 21:25 |
zer0def | well, there's just a cascade of things going wrong | 21:25 |
kata-irc-bot | <archana.m.shinde> zer0def ^ | 21:25 |
zer0def | yeah, I got that, would still feel more comfy about it *after* drilling down into why conmon returns -1 instead of container PID that then leads podman to make a bad choice | 21:26 |
zer0def | and since conmon basically calls the runtime, I figured it might be worthwhile to bring up | 21:26 |
kata-irc-bot | <archana.m.shinde> yeah, could be a runtime issue, cant rule that out | 21:29 |
kata-irc-bot | <archana.m.shinde> let me see | 21:29 |
zer0def | that said, it's been a week or more since I've dug into this and my memory's a little fuzzy, since I was doing it right before sleep | 21:31 |
kata-irc-bot | <eric.ernst> :slightly_smiling_face: | 21:31 |
zer0def | my hypothesis would be that both issues may have a common root in kata-agent misbehaving somehow, leading to various cascades of errors | 21:33 |
*** sameo has joined #kata-general | 22:34 | |
*** sameo has quit IRC | 23:06 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!