Thursday, 2019-06-20

*** tmhoang has quit IRC02:04
*** sameo_ has joined #kata-general04:34
*** tmhoang has joined #kata-general05:56
*** tmhoang has quit IRC05:59
*** lpetrut has joined #kata-general06:06
*** tmhoang has joined #kata-general06:06
*** sameo_ has quit IRC06:45
*** sameo_ has joined #kata-general07:43
*** gwhaley has joined #kata-general08:03
*** igordc has quit IRC08:20
*** gwhaley has quit IRC10:58
*** lpetrut has quit IRC11:35
*** gwhaley has joined #kata-general12:31
kata-irc-bot<graham.whaley> OK... @gmmaharaj, yeah, looks like there are some niggles in the data grabbing when we have 'non-standard' kata env setups - that is, when we moved to static and kata-deploy and nemu style installations with different names/paths, some things we have to assume broke :( So, the files got a broken json kata-env section in them, which stopped them being processed. I just nuked the broken entries and processed (after fixing another12:51
kata-irc-botsmall niggle with the report generator). Let's see if I can attach the pdf here... slack is looking like it will take the drag/drop, but I'm not convinced.... if not, I'll try to paste images (4 pages)... overall, looks like virtiofs is a win on reads, not so much on writes - that is probably due to the cache/writeback-through model used in the tests is my guess (iirc 9p caches a lot in the guest on writes for instance).12:51
kata-irc-bot<graham.whaley> metrics comparison of 9p vs virtiofs in fio tests - pages 7-1012:52
kata-irc-bot<graham.whaley> Pages 7-10 are the ones you want. the rest are empty as we only ran the fio data grab. Heh, err, if anybody wants to help me make that data any better, I think I had some ideas about trying to improve that latency percentile over block graph - thought I had an Issue open, but apparently not...12:55
*** sameo_ has quit IRC15:53
*** EricAdamsZNC2 has joined #kata-general15:59
*** EricAdamsZNC has quit IRC16:03
*** sameo has joined #kata-general16:28
*** gwhaley has quit IRC16:52
*** tmhoang_ has joined #kata-general18:08
*** tmhoang has quit IRC18:09
*** igordc has joined #kata-general18:27
*** tmhoang_ is now known as tmhoang18:28
*** igordc has quit IRC18:36
*** sameo has quit IRC20:42

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!