[09:47:20] <nemo-yiannis>	 👋 We are in the process of migrating rt-testing infra from an internal vm to cloudvps. This isn't production facing infrastructure but we heavily rely on this for our weekly release confidence for parsoid. Currently we have a VM with mysql data stored in a volume. Would this make any difference in performance if we use a database instance (trove)?
[09:48:37] <volans>	 nemo-yiannis: how much data are we talking about? Have you considered toolforge instead?
[10:12:34] <nemo-yiannis>	 For now, it better fits our workflow/scripts/etc a simple VM with SSH access
[10:13:12] <nemo-yiannis>	 Our current DB is ~30G 
[10:13:28] <nemo-yiannis>	 but this grows depending on the runs to max ~100G
[10:14:37] <nemo-yiannis>	 We cleanup eventually old runs that are not needed for our investigations
[10:17:41] <nemo-yiannis>	 Eventually (in the future) we are going to move to a more distributed setup that could fit toolforge but currently we need to have a single dedicated VM
[10:27:54] <volans>	 nemo-yiannis: ack, trove should be fine, in terms of performance is equivalent of having the DB in another VM
[10:28:35] <nemo-yiannis>	 so no difference between an attached volume to a VM and a trove instance ?
[10:28:54] <nemo-yiannis>	 then we are probably good with our current setup
[10:30:47] <volans>	 it's more standard, easier to manage given the abstraction around it
[10:34:38] <nemo-yiannis>	 sounds good, thanks!
[10:36:54] <volans>	 but yeah, not too far from your existing setup
[14:55:54] <andrewbogott>	 !log admin "systemctl restart ceph-osd@289.service" as 289 shows as down
[14:55:59] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL
[14:57:57] <andrewbogott>	 !log admin "systemctl restart ceph-osd@281.service" as 281 shows as down
[14:57:59] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL
[15:13:48] <andrewbogott>	 !log admin systemctl restart ceph-osd@58.service due to slow ops
[15:13:54] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL
[15:15:19] <dcausse>	 !log toolhub drop toolhub_* indices from the main cirrussearch opensearch cluster (T423993)
[15:15:21] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolhub/SAL
[15:15:22] <stashbot>	 T423993: Upgrade old indices in the CirrusSearch opensearch clusters - https://phabricator.wikimedia.org/T423993
[15:17:19] <dcausse>	 ok updated https://docs.google.com/spreadsheets/d/1MYH3SScPgwPUoQv8gTTywOegTcrj-W33uA0Bw1gD3yg/edit?gid=2121474748#gid=2121474748 (only ttmserver AFAICT)
[15:17:45] <andrewbogott>	 !log sudo systemctl restart ceph-osd@61.service
[15:17:45] <stashbot>	 andrewbogott: Unknown project "sudo"
[15:33:00] <andrewbogott>	 !log admin restarting quite a few other ceph-osd services in an attempt to quiet some slow op alerts
[15:33:06] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL
[17:04:47] <andrewbogott>	 !log rebooting cloudcephosd1037 and cloudcephosd1038 because of missing volumes. The volumes reappeared after boot.
[17:04:48] <stashbot>	 andrewbogott: Unknown project "rebooting"
[17:04:57] <andrewbogott>	 !log admin rebooting cloudcephosd1037 and cloudcephosd1038 because of missing volumes. The volumes reappeared after boot.
[17:05:01] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL
[17:51:27] <andrewbogott>	 !log admin ceph tell osd.126,127,129,131  compact
[17:51:33] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL