Closed jimklimov closed 4 years ago
The report (as seen in Travis at least) resulted in a strange markup:
[Thu Oct 24 12:05:38 2019] [warn] ERROR: suspending cleanup source dataset because 3 send task(s) failed:\ndestination 'backup/tank/anothersource' does not exist or is offline. ignoring it for this round...\n\tdestination 'root@remote:remote/tank/anothersource' does not exist or is offline. ignoring it for this round...\n\tdestination 'backup/destfail' does not exist or is offline. ignoring it for this round...\n
probably some qw
inside? Should I change this to a loop of warn()
calls for each array element?
Ok, with a loop it looks better:
[Thu Oct 24 12:54:26 2019] [warn] ERROR: suspending cleanup source dataset because 3 send task(s) failed:
[Thu Oct 24 12:54:26 2019] [warn] +--> destination 'backup/tank/anothersource' does not exist or is offline. ignoring it for this round...
[Thu Oct 24 12:54:26 2019] [warn] +--> destination 'root@remote:remote/tank/anothersource' does not exist or is offline. ignoring it for this round...
[Thu Oct 24 12:54:26 2019] [warn] +--> destination 'backup/destfail' does not exist or is offline. ignoring it for this round...
[Thu Oct 24 12:54:26 2019] [info] done with backupset tank/anothersource in 1 seconds
or (hey, typo!)
[Thu Oct 24 12:54:41 2019] [warn] ERROR: suspending cleanup source dataset because 2 send task(s) failed:
[Thu Oct 24 12:54:41 2019] [warn] +--> skipping backup/destinationdue to pre-command failure
[Thu Oct 24 12:54:41 2019] [warn] +--> destination 'backup/destfail' does not exist or is offline. ignoring it for this round...
or
[Thu Oct 24 12:55:01 2019] [warn] ERROR: suspending cleanup source dataset because 2 send task(s) failed:
[Thu Oct 24 12:55:01 2019] [warn] +--> ERROR: cannot send snapshots to backup/destination
[Thu Oct 24 12:55:01 2019] [warn] +--> destination 'backup/destfail' does not exist or is offline. ignoring it for this round...
[Thu Oct 24 12:55:01 2019] [info] done with backupset tank/source in 3 seconds
Cleaned up (hopefully) the rebase of many masters and PR iterations of the day... now.
I guess soon it will be hard for PRs to not break coverage percentage :-)
thanks
Welcome! :)
Solution: repeat the warnings/errors logged when flagging a send as failed in the end, to help find the needles in the haystask. Reuse the previously booleanish flag $sendFailed as an array of such messages (usually one line = one sending error) @sendFailed.
Signed-off-by: Jim Klimov jimklimov@gmail.com