[16:03:12] <cdanis>	 brt
[16:21:24] <jijiki>	 bart?
[16:57:36] <jijiki>	 hey all, anyone knows what we should do with the swift server alerts?
[16:58:37] <jijiki>	 I logged into a couple of them 
[16:58:43] <jijiki>	 and I see messages like 
[16:58:45] <jijiki>	 task xfs_db:11362 blocked for more than 120 seconds.
[16:58:58] <jijiki>	 task kswapd0:319 blocked for more than 120 seconds.
[16:59:15] <jijiki>	 etc in dmesg
[17:15:28] <cdanis>	 which machines are those on jijiki
[17:15:30] <cdanis>	 ?
[17:16:00] <jijiki>	 I checked ms-be2028 a few minutes ago
[17:16:20] <jijiki>	 and a couple of hours ago I checkd ms-be1031
[17:16:26] <cdanis>	 I don't know how common it is overall, but those hosts are probably very busy right now, moving a bunch of data around for decomming a few machines
[17:17:16] <jijiki>	 same for ms-be2026
[17:17:22] <jijiki>	 oh I see
[17:17:40] <jijiki>	 it is not common as the rest of dmesg looks normal 
[17:17:47] <jijiki>	 tx cdanis 
[17:18:05] <cdanis>	 it's still not great we're getting such errors, and that NRPEs are timing out
[17:19:39] <cdanis>	 I don't think Swift has any support for setting ratelimits on replication traffic though, maybe godog knows better
[17:41:48] <godog>	 yeah not swift itself the software but we could limit rsync
[23:40:34] <quiddity>	 (brt = be right there :)