| noonedeadpunk | clarkb: yup, that was me with u-c fetch issues. was having really bad times with gitea yesterday during EU daytime overall | 11:30 |
|---|---|---|
| Clark[m] | noonedeadpunk that was an independent issue (seems we were overwhelmed with AI web crawling that we think we mitigated). The upgrade separately should fix a memory leak that I hope will result in better reliability overall | 16:03 |
| clarkb | I don't see any email complaints about failed backups which is a good sign. I can't login and actually check backup logs or archives directly this morning though | 16:56 |
| fungi | Backups failed on host paste02 at Sat Jan 18 05:29:35 UTC 2025. | 19:49 |
| fungi | Backups failed on host paste02 at Sat Jan 18 17:29:30 UTC 2025. | 19:49 |
| fungi | rackspace mail apparently decided those were spam, which is why they weren't in our inbox | 19:51 |
| fungi | i have moved them from the spam folder to the inbox now | 19:51 |
| fungi | Warning: Attempting to access a previously unknown unencrypted repository! | 19:57 |
| fungi | Do you want to continue? [yN] 1 (from BORG_UNKNOWN_UNENCRYPTED_REPO_ACCESS_IS_OK) | 19:57 |
| fungi | Creating archive at "ssh://borg-paste02@backup01.ord.rax.opendev.org/opt/backups/borg-paste02/backup::paste02-filesystem-2025-01-18T17:26:02" | 19:58 |
| fungi | [...] | 19:58 |
| fungi | terminating with warning status, rc 1 | 19:58 |
| fungi | [...] | 19:58 |
| fungi | Sat Jan 18 17:29:30 UTC 2025 Backup finished with errors | 19:59 |
| fungi | i suspect it's the warning causing the nonzero exit, i didn't see any other obvious explanations | 19:59 |
| fungi | also that didn't seem to prevent it from proceeding with the backup | 20:00 |
| fungi | seems like that's normal, from what i'm reading, but may not be able to silence it if we do unencrypted... we might need to work out how to accept rc 1? | 20:10 |
| fungi | anyway, i'm heading to dinner, this seems non-urgent, if noisy | 20:10 |
| Clark[m] | fungi: ya rc1 means exited with warning(s) likely the same file changed while backing it up we saw in CI. I agree non urgent. We can probably update the script to log rc 1 but then change it to a 0 if 0 or 1 | 20:42 |
| Clark[m] | Problem is the old system we should still error on 1 so maybe needs more smarts | 20:42 |
| fungi | right, i figure we need to mull it over and discuss options | 21:49 |
| Clark[m] | https://news.ycombinator.com/item?id=42750420 | 23:43 |
| Clark[m] | The discussion there seems relevant to our experience yesterday. Lots of info on who does and who doesn't respect which robots.txt directives and other info. Granted not sure how accurate it is | 23:44 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!