Tuesday, 2020-03-31

kota_morning00:43
kota_timburke: good, you knew that.00:43
kota_it looks SFO has been under shelter-in-place order tho...00:46
*** diablo_rojo has quit IRC02:24
*** diablo_rojo has joined #openstack-swift02:24
*** psachin has joined #openstack-swift03:29
*** threestrands has joined #openstack-swift03:39
*** ccamacho has quit IRC04:08
*** evrardjp has quit IRC04:36
*** evrardjp has joined #openstack-swift04:36
*** diablo_rojo has quit IRC04:53
*** TerianceVi has joined #openstack-swift05:04
*** TViernion has quit IRC05:05
*** rpittau|afk is now known as rpittau07:17
*** diablo_rojo has joined #openstack-swift07:32
*** threestrands has quit IRC08:26
*** rcernin has quit IRC08:54
*** diablo_rojo has quit IRC09:37
*** rcernin has joined #openstack-swift10:00
*** ccamacho has joined #openstack-swift10:07
*** rcernin has quit IRC10:14
*** rpittau is now known as rpittau|bbl11:05
*** psachin has quit IRC11:35
*** psachin has joined #openstack-swift11:36
*** psachin has quit IRC11:53
*** psachin has joined #openstack-swift11:55
*** rpittau|bbl is now known as rpittau12:15
*** tkajinam has quit IRC14:45
timburkegood morning15:05
*** diablo_rojo has joined #openstack-swift15:15
*** diablo_rojo has quit IRC15:50
*** rpittau is now known as rpittau|afk16:07
timburkerledisez, on p 715577, i'm kind of surprised that the diskfile operations in _check_local() wouldn't trip a trampoline and let other greenthreads run -- are those all blocking?16:22
patchbothttps://review.opendev.org/#/c/715577/ - swift - Fix SSYNC/missing_check is blocking - 2 patch sets16:22
*** evrardjp has quit IRC16:36
*** evrardjp has joined #openstack-swift16:36
*** psachin has quit IRC16:50
openstackgerritRomain LE DISEZ proposed openstack/swift feature/losf: Add missing option in for the volume check script  https://review.opendev.org/71391518:24
openstackgerritRomain LE DISEZ proposed openstack/swift feature/losf: Remove migrator.py  https://review.opendev.org/71391618:24
rlediseztimburke: at first I though that "with exceptions.MessageTimeout" or "self.fp.readline()" would allow gth switching, but it didn't. I falied to understand why so I just added the sleep(), but from the diskfile… I though that disk I/O was still blocking with eventlet? (reason of tpool execute & co)18:36
*** diablo_rojo has joined #openstack-swift18:56
timburkerledisez, i know the readline() *can* allow a trampoline, but only if the read would block. timeouts won't trampoline, by they may raise during a trampoline19:04
timburke*shrug* at the very least, the occasional extra sleep is pretty low-risk19:04
openstackgerritRomain LE DISEZ proposed openstack/swift master: Make wsgi server uses systemd's NOTIFY_SOCKET  https://review.opendev.org/71557619:22
*** rcernin has joined #openstack-swift21:02
openstackgerritRomain LE DISEZ proposed openstack/swift master: relinker: Improve performance by limiting I/O  https://review.opendev.org/69534421:33
mattoliverauMorning22:42
*** tkajinam has joined #openstack-swift22:49
*** ianychoi has quit IRC23:15
*** ianychoi has joined #openstack-swift23:16
openstackgerritMerged openstack/swift master: Fix SSYNC/missing_check is blocking  https://review.opendev.org/71557723:43
openstackgerritMerged openstack/swift feature/losf: Add missing option in for the volume check script  https://review.opendev.org/71391523:43
openstackgerritMerged openstack/swift feature/losf: Remove migrator.py  https://review.opendev.org/71391623:43

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