[ For future reference, -general is the appropriate list. Moving
discussion there. ]
On Sat, 2011-06-04 at 00:45 +0300, Alexander Shulgin wrote:
> We've noticed that free disk space went down heavily on a system, and
> after a short analysis determined that the reason was that postmaster
> was holding lots of unlinked files open. A sample of lsof output was
> something like this:
...
> Restarting PostgreSQL obviously helps the issue and the disk space
> occupied by those unlinked files (about 63GB actually) is reclaimed.
Normally postgres closes unlinked files during a checkpoint. How long
between checkpoints on this system? Is it possible that you noticed
before postgresql caused an automatic checkpoint?
Also, you can do a manual checkpoint with the CHECKPOINT command.
Regards,
Jeff Davis