Friday, 2018-10-26

openstackgerritTim Burke proposed openstack/swift master: WIP: s3api: Move authenticator logic to separate module  https://review.openstack.org/61345200:26
*** itlinux has joined #openstack-swift00:49
*** patchbot has quit IRC00:57
*** patchbot has joined #openstack-swift00:57
*** two_tired has joined #openstack-swift01:28
*** rcernin has joined #openstack-swift01:55
*** two_tired has quit IRC01:56
kota_mattoliverau: morning03:28
mattoliveraukota_: o/03:32
mattoliverauSo I've been trying to recreate bug 1781291, but attempting on a virtual cluster on my dev laptop. I can get it to pike in container GET latency.. but not really huge. I've gone an added an index to the shard_range table and yeah, maybe an improvement. But I think my problem is simulatiing on a cluster sharing the same hardware makes it a little harder to recreate, as it's all off the same physical disk.03:39
openstackbug 1781291 in OpenStack Object Storage (swift) "sharding: container GETs to root container get slow" [Medium,New] https://launchpad.net/bugs/178129103:39
mattoliverauI try and throw more at it, and test and see if I can even get slight iops differences.03:39
mattoliveraumaybe I'll see if I can move this to the engineering cloud at work. Still use VMs but more of them which will hopeully be on different boxes and see if that goes better.03:40
mattoliverauEither that or try and talk the wife into letting build a cluster at home ;)03:40
mattoliveraua _real_ cluster03:41
notmynamemattoliverau: something else to look for is the container sharder consuming a *lot* of CPU when there isn't anything to do. it's something we've seen a few times. not sure if it's merely a config thing or something actually that should be changed in code03:48
notmyname(we've been "fixing" it by changing the interval)03:48
mattoliverauahh good to know, I can look for that on this little cluster03:49
mattoliveraualso a good chance to do some sharder profiling03:50
openstackgerritPete Zaitcev proposed openstack/swift master: py3: adapt the account server completely  https://review.openstack.org/61350504:44
openstackgerritPete Zaitcev proposed openstack/swift master: Fix up the test for .ismount  https://review.openstack.org/61350605:29
openstackgerritPete Zaitcev proposed openstack/swift master: Fix up the test for .ismount  https://review.openstack.org/61350605:31
*** mahatic_ has joined #openstack-swift05:48
*** ChanServ sets mode: +v mahatic_05:48
*** cwright_ has joined #openstack-swift05:51
*** mahatic has quit IRC05:55
*** irclogbot_2 has quit IRC05:55
*** mattoliverau has quit IRC05:55
*** cwright has quit IRC05:55
*** jlvillal has quit IRC05:55
*** aluria has quit IRC05:55
*** hugokuo has quit IRC05:55
*** ccamacho has quit IRC06:20
*** ccamacho has joined #openstack-swift06:41
*** ccamacho has quit IRC06:49
*** ccamacho has joined #openstack-swift06:51
*** rcernin has quit IRC07:22
openstackgerritAlexandre Lécuyer proposed openstack/swift master: Remove empty directories after a revert job  https://review.openstack.org/61132507:36
*** mikecmpbll has joined #openstack-swift07:56
*** ccamacho has quit IRC08:32
*** ccamacho has joined #openstack-swift08:33
openstackgerritRomain LE DISEZ proposed openstack/swift master: Make log format for requests configurable  https://review.openstack.org/54894809:03
*** e0ne has joined #openstack-swift09:16
*** e0ne has quit IRC10:51
*** e0ne_ has joined #openstack-swift10:52
*** e0ne_ has quit IRC12:26
*** e0ne has joined #openstack-swift13:25
*** cwright_ is now known as cwright13:32
*** onovy has quit IRC14:16
*** itlinux has quit IRC15:13
*** zaitcev has quit IRC15:13
*** gyee has joined #openstack-swift15:23
*** zaitcev has joined #openstack-swift15:23
*** ChanServ sets mode: +v zaitcev15:23
*** onovy has joined #openstack-swift15:31
*** itlinux has joined #openstack-swift15:56
*** mikecmpbll has quit IRC16:02
*** e0ne has quit IRC16:10
notmynamegood morning16:55
notmynamewe don't seem to be affected by this, but please be aware of http://lists.openstack.org/pipermail/openstack-dev/2018-October/136099.html17:01
notmynamezaneb will be attending next week's swift team meeting where we will discuss https://review.openstack.org/#/c/592205/17:20
patchbotpatch 592205 - governance - [DRAFT] Add a Technical Vision statement - 4 patch sets17:20
notmynameah, he may not make it, in fact. either way, we'll still discuss the patch17:22
zaitcevI started looking at that patch but was neck deep in py3 and forgot17:52
zaitcevI mean the "vision"17:52
timburkezaitcev: you made me think of https://www.youtube.com/watch?v=7xPBHn7Cblk&t=7717:59
timburke:P17:59
zaitcevWell, I'm basically done with the auditor18:01
zaitcever.18:01
zaitcevI meant to say, with the account server, including auditor, replicator, server.18:01
zaitcevBUT18:01
zaitcevMy UTF-8 encoded query parameters made a repeat appearance... So, I'll be relying on you to give it another look, in view of your ability to simplify my excessively elaborate approaches.18:02
openstackgerritPete Zaitcev proposed openstack/swift master: py3: adapt the account server completely  https://review.openstack.org/61350518:03
zaitcevtimburke: BTW, please take a look at https://review.openstack.org/#/c/613506/ , it's very small18:05
patchbotpatch 613506 - swift - Fix up the test for .ismount - 2 patch sets18:05
*** e0ne has joined #openstack-swift18:35
*** e0ne has quit IRC19:14
*** e0ne has joined #openstack-swift19:19
*** e0ne has quit IRC19:23
openstackgerritTim Burke proposed openstack/swift master: WIP: s3api: Move authenticator logic to separate module  https://review.openstack.org/61345219:23
*** openstackstatus has quit IRC20:42
*** openstack has joined #openstack-swift20:45
*** ChanServ sets mode: +o openstack20:45
*** mikecmpbll has joined #openstack-swift20:55
*** mikecmpbll has quit IRC21:41
*** itlinux has quit IRC22:05
*** pcaruana has quit IRC23:10
*** rcernin has joined #openstack-swift23:44
*** rcernin has quit IRC23:51
*** itlinux has joined #openstack-swift23:52
*** gyee has quit IRC23:55

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