Pitrery wal log backup after failure about pitrery HOT 2 CLOSED

Hello,

Since it PostgreSQL that runs the archive_command, it does not try to archive the next file until the previous archive_command execution has succeeded. So I wonder how this could happen in the first place.

The backup will not re-archive, it is PostgreSQL that takes care of archiving, pitrery just wait for PostgreSQL to let it know every WAL file needed for a consistent recovery are archived. So taking a new backup and copying the missing files is a good way to go (use pitrery check -A to have pitrery test if there are not any missing archived WAL files)

Concerning the backup then transfer idea, I guess the attached storage has to be reliably attached at some point somewhere, for the transfer to succeed. So I would work on making it more reliable in the first place before complicating things. If your storage has availability issues, I would suggest to enable integrity options of archive_xlog (-H or ARCHIVE_CHECK="yes" and -F or ARCHIVE_FLUSH="yes")

Regards

from pitrery.

Fantastic. Thank you for the help. This actually happened for the first time, so I just prepared in case if happens more times. I'll update the parameters just in case.

As a side note, do you plan to put any notifications of failed backups (either base or wal) in the script?

Thanks.
-N.

from pitrery.