Thanks for the info on alarm and timeouts, this was a big help. One further comment:
Michael Fuhr wrote:
eval {
local $SIG{ALRM} = sub {die("Timeout");};
$time = gettimeofday;
alarm 20;
$sth = $dbh->prepare("a query that may take a long time...");
$sth->execute();
alarm 0;
};
if ($@ && $@ =~ /Timeout/) {
my $elapsed = gettimeofday - $time;
print "Timed out after $elapsed seconds";
}
Now the mystery: It works, but it hardly matters what time I use for the
alarm call, the actual alarm event always happens at 26 seconds...
High-level languages' signal handlers don't always work well with
low-level libraries...
Is there a reason you're using alarm() in the client instead of
setting statement_timeout on the server?
statement_timeout solved the problem, thanks VERY much for the pointer. To answer your question, I use alarm() because all the books and web references said that was how to do it. :-). I've used alarm() with Perl (with a 3rd-party C lib that had a potential infinite loop) very successfully.
So thanks for the pointer to statement_timeout. But...
When I set statement_timeout in the config file, it just didn't do anything - it never timed out (PG 8.0.3). I finally found in the documentation that I can do "set statement_timeout = xxx" from PerlDBI on a per-client basis, and that works.
Craig
---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?
http://archives.postgresql.org
