| *** lcastell has joined #kata-general | 00:02 | |
| *** lcastell has quit IRC | 00:06 | |
| kata-irc-bot | <raravena80> :+1: | 02:15 |
|---|---|---|
| *** sameo has joined #kata-general | 05:02 | |
| *** igordc has quit IRC | 05:19 | |
| *** kata-irc-bot has quit IRC | 05:45 | |
| *** kata-irc-bot has joined #kata-general | 05:45 | |
| *** tmhoang has joined #kata-general | 06:12 | |
| kata-irc-bot | <voje> @eric.ernst thank you, this presentation is excellent! Exactly what I was looking for. | 06:14 |
| *** zamzing has quit IRC | 06:31 | |
| *** sgarzare has joined #kata-general | 06:42 | |
| *** sameo has quit IRC | 06:48 | |
| *** auk has quit IRC | 07:06 | |
| *** sameo has joined #kata-general | 07:42 | |
| *** gwhaley has joined #kata-general | 08:04 | |
| kata-irc-bot | <graham.whaley> @david532 - the kata runtime also does memory hotplug into the qemu. Thus, for k8s pods for instance, if more containers are added that are requesting more memory resource (in their yaml definitions), then kata will hotplug in more memory (via qemu memory dimm slot hotplug). note, we don't hot-unplug if/when those containers then later go away, as that is much more expensive and difficult. | 08:27 |
| *** sgarzare has quit IRC | 15:44 | |
| *** gwhaley has quit IRC | 16:50 | |
| *** igordc has joined #kata-general | 17:28 | |
| kata-irc-bot | <david532> Thank you! I know memory can be a bit of a sticky situation, definitely in these cases. I suppose something in the terms of an overcommit on the host should assist with density. | 17:53 |
| *** igordc has quit IRC | 19:44 | |
| *** igordc has joined #kata-general | 19:46 | |
| *** sameo has quit IRC | 20:09 | |
| *** tmhoang has quit IRC | 20:29 | |
| *** lcastell has joined #kata-general | 20:36 | |
| *** lcastell has quit IRC | 22:01 | |
| *** auk has joined #kata-general | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!