Re: Something else about Redo Logs disappearing

On Mon, Jun 15, 2020 at 03:19:29PM +0200, Laurenz Albe wrote:
! On Mon, 2020-06-15 at 14:50 +0200, Peter wrote:
! > ! An example:
! > ! 
! > ! - Backup #1 calls "pgpre.sh"
! > ! - Backup #1 starts copying files
! > ! - Backup #2 calls "pgpre.sh".
! > !   This will cancel the first backup.
! > ! - Backup #1 completes copying files.
! > ! - Backup #1 calls "pgpost.sh".
! > !   It will receive an error.
! > !   So it has to invalidate the backup.
! > ! - Backup #2 completes copying files.
! > ! - Backup #2 calls "pgpost.sh".
! > !   It gets a "backup_label" file and completes the backup.
! > 
! > That's not true.
! 
! Ah, yes, you are right.

Thank You.

! Since "pgpre.sh" and "pgpost.sh" are independent, there
! is no way to tell which of them belongs to which other.

Correct.
 
! So calling "pgpost.sh" indeed ends the most recently started
! backup and returns "backup_label" accordingly.
! 
! That means: the caller of the scripts has to make sure
! not to start a second backup while the first one is running.

Never run two backups in parallel with such an approach, exactly.

And that is one of a couple of likely pitfalls I perceived when
looking at that new API.

We could fix that, but that will then get more complicated - and
people will usually not do that. And that's why I consider that
new API as rather dangerous.


cheerio,
PMc