| jcrespo | |
| Oct 10 2024, 12:12 PM |
| F57744915: image.png | |
| Nov 25 2024, 9:02 AM |
| F57744909: image.png | |
| Nov 25 2024, 9:02 AM |
| F57604423: image.png | |
| Oct 10 2024, 12:25 PM |
We recently run into low space issues, expand the dedicated size for media backups to 960 TB (6 hosts) on both eqiad and codfw.
Current utilization (as of 2024-10-10) is 645 708 607 840 309 bytes.
Current utilization (as of 2024-11-21) is 656 178 463 691 108 bytes.
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Open | None | T262668 WMF media storage must be adequately backed up | |||
| Resolved | jcrespo | T376892 Expand media backup storage available space to 960 TB per datacenter | |||
| Resolved | Jclark-ctr | T371416 Q1:rack/setup/install backup1012 | |||
| Resolved | Jhancock.wm | T371984 Q1:rack/setup/install backup2012 |
jynus opened https://gitlab.wikimedia.org/repos/sre/mediabackups/-/merge_requests/4
mediabackups: Prepare reshard for backup*012, new host
jynus closed https://gitlab.wikimedia.org/repos/sre/mediabackups/-/merge_requests/4
mediabackups: Prepare reshard for backup*012, new host
jynus opened https://gitlab.wikimedia.org/repos/sre/mediabackups/-/merge_requests/5
mediabackups: Prepare reshard for backup*012, new host
Change #1082172 had a related patch set uploaded (by Jcrespo; author: Jcrespo):
[operations/puppet@production] mediabackups: Setup new host for mediabackups backup[12]012
Change #1082172 merged by Jcrespo:
[operations/puppet@production] mediabackups: Setup new host for mediabackups backup1012
Change #1091731 had a related patch set uploaded (by Jcrespo; author: Jcrespo):
[operations/puppet@production] backup: Move Dell bacula hosts to mediabackups
Change #1091731 merged by Jcrespo:
[operations/puppet@production] backup: Move Dell bacula hosts to mediabackups
Mentioned in SAL (#wikimedia-operations) [2024-11-20T14:23:26Z] <jynus> starting resharding of commons backup files into new host backup1010 T376892
Mentioned in SAL (#wikimedia-operations) [2024-11-20T15:31:53Z] <jynus> starting resharding of commons backup files into new host backup2010 T376892
Change #1093377 had a related patch set uploaded (by Jcrespo; author: Jcrespo):
[operations/puppet@production] mediabackup: Setup backup1010 as the 6th media backup host in eqiad
Change #1093379 had a related patch set uploaded (by Jcrespo; author: Jcrespo):
[operations/puppet@production] mediabackup: Setup backup2010 as the 6th media backup host in codfw
Change #1093377 merged by Jcrespo:
[operations/puppet@production] mediabackup: Setup backup1010 as the 6th media backup host in eqiad
Change #1093379 merged by Jcrespo:
[operations/puppet@production] mediabackup: Setup backup2010 as the 6th media backup host in codfw
Timestamp is in CET:
[12:26:36] <jinxer-wm> RESOLVED: DiskSpace: Disk space backup2011:9100:/srv/objectstorage 5.998% free - https://wikitech.wikimedia.org/wiki/Monitoring/Disk_space - https://grafana.wikimedia.org/d/000000377/host-overview?orgId=1&viewPanel=12&var-server=backup2011 - https://alerts.wikimedia.org/?q=alertname%3DDiskSpace
This is now done. Catchup and purging is ongoing, but after that finishes, we should be able to store almost 1 PB of media backups on both datacenters. Backups will continue without problem or manual interventions in the next months.
Data is not 100% balanced, so that may require later tunings, but the initial goal, which was expanding the existing space and make sure hosts were not full, was accomplished already.
