Temporary sorting space

On Wed, Mar 07, 2001 at 04:11:23PM +1300, Richard Boyes wrote:

Hello all,

Inside the $PGDATA/base/thedatabase I came across a rather
large file called pg_sorttemp28193.0 (~ 587 MB). Our datbase
is not anywhere near this large, so I can only guess this was
a query gone awry that needed a large amount of sorting space.
Around the time this file was created the database server
also crashed, I'm wondering if this was a cause of it
(the server crashed due to a memory problem). Could that be possible ?

Also, would I now be safe in removing this file ?

i'm not much of a guru, but here's how i'd test the safety of
removing any file:

rename it or put it into a
subdirectory of wherever it is now.

then restart your pg engine and see if something breaks. test
heavily. after a backup, and a week of no troubles, zap your file.
if something breaks, put it back!

--
It is always hazardous to ask "Why?" in science, but it is often
interesting to do so just the same.
-- Isaac Asimov, 'The Genetic Code'

will@serensoft.com
http://newbieDoc.sourceforge.net/ -- we need your brain!
http://www.dontUthink.com/ -- your brain needs us!

Richard Boyes <r.boyes@auckland.ac.nz> writes:

Inside the $PGDATA/base/thedatabase I came across a rather
large file called pg_sorttemp28193.0 (~ 587 MB). Our datbase
is not anywhere near this large, so I can only guess this was
a query gone awry that needed a large amount of sorting space.
Around the time this file was created the database server
also crashed, I'm wondering if this was a cause of it
(the server crashed due to a memory problem). Could that be possible ?

Possibly, but this is not a lot of evidence. Can you provide any
log file entries, core dump backtraces, etc?

Also, would I now be safe in removing this file ?

If there's not currently a backend with PID 28193, then that's a
dead file. Zap away.

regards, tom lane