| *** bpradipt has joined #kata-general | 04:21 | |
| *** sgarzare has joined #kata-general | 06:36 | |
| *** sameo has quit IRC | 09:28 | |
| *** pvdp66556 has joined #kata-general | 09:42 | |
| *** sameo has joined #kata-general | 09:44 | |
| *** kata-irc-bot has quit IRC | 15:36 | |
| *** kata-irc-bot has joined #kata-general | 15:36 | |
| *** kata-irc-bot has quit IRC | 15:36 | |
| *** kata-irc-bot has joined #kata-general | 15:37 | |
| *** sgarzare has quit IRC | 16:02 | |
| *** tobberydberg has quit IRC | 16:51 | |
| *** tobberydberg_ has joined #kata-general | 16:51 | |
| *** bpradipt has quit IRC | 17:57 | |
| kata-irc-bot | <simon.kaegi> That looks good -- so now you're just looking at the overhead associated with virtio-fs on whatever your underlying storage is. Devmapper will perform better in many cases as virtio-fs is still actively being improved. | 17:59 |
|---|---|---|
| kata-irc-bot | <simon.kaegi> On bare metal nodes mounting local disk storage for our ci/cd workloads we see a penalty in the range of 20%. When operating through two layers of virtualized disk we've seen the penalty to be much higher. Definitely would suggest using block storage in those cases | 18:01 |
| *** sameo has quit IRC | 23:02 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!