"What do I do after a deep-check fails?"

Getting Help
1

Running /usr/bin/restic -r r check --read-data as a systemd-service gives me:

restic[364602]: [14:57:42] 100.00%  77351 / 77351 packs
restic[364602]: Fatal: repository contains errors
systemd[5819]: restic-check-deep.service: Main process exited, code=exited, status=1/FAILURE

However … what do I do now? There is no “leading text”, given that this is a systemd service (and that has been long rotated by now) :confused:

Documentation offers no walkthrough on what to do :confused:

2

Might be worth trying with a -v or two:

[david@pc ~]$ restic check --help | grep verbose
  -v, --verbose                    be verbose (specify multiple times or a level using --verbose=n, max level/times is 2)
[david@pc ~]$
3

I think the issue was not well-explained; apologies.

systemd writes to journalctl; status is only a tail of the (rolling) output.
Additionally, “under some conditions”, journalctl can rotate the output to /dev/null - hence, no output

Adding more verbose output won’t help in this situation :confused:

4

It’s pretty unusual for journald to drop the logs within a few hours (did you check the output using journalctl or systemctl?). As we’re missing the relevant log output here, you have to either fix the log retention or modify the service to redirect the log output to a separate logfile that won’t be removed too early.

That sounds like systemctl status service-name. Please use journalctl -u service-name instead.

That won’t make a difference. Relevant errors are always printed independent of the current verbosity level.

5

So … there was output somewhere that I missed it? :confused:

Output got thrown out due to some dmesg issues being overly spammy (and right now I have very poor internet to retry this).

I’ll retry this later :upside_down_face:

6

Yes, the details for a warning / error is printed immediately when restic detects it.

1 Like