Re: Unresolved error 0xC0000409 on Windows Server

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Matthew Gerber <gerber(dot)matthew(at)gmail(dot)com>
Cc: Noah Misch <noah(at)leadboat(dot)com>, Craig Ringer <ringerc(at)ringerc(dot)id(dot)au>, Merlin Moncure <mmoncure(at)gmail(dot)com>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: Unresolved error 0xC0000409 on Windows Server
Date: 2012-11-11 17:22:24
Message-ID: 24361.1352654544@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Matthew Gerber <gerber(dot)matthew(at)gmail(dot)com> writes:
> On Sun, Nov 11, 2012 at 11:19 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> How long is "long"?

> I was seeing queries with around 5000-7000 "UNION ALL" statements.

Hm. I experimented with test queries created like so:

perl -e 'print "SELECT 1 a, 2 b, 3 c\n"; print "UNION ALL SELECT 1 a, 2 b, 3 c\n" foreach (1..8200);' | psql

On the machine I tried this on, it works up to about 8200 and then fails
in the way I'd expect:

ERROR: stack depth limit exceeded
HINT: Increase the configuration parameter "max_stack_depth" (currently 2048kB), after ensuring the platform's stack depth limit is adequate.

But then when I cranked it up to 80000, kaboom:

connection to server was lost

Inspection of the core dump shows transformSetOperationTree is the
problem --- it's recursing but lacks a check_stack_depth test.
So that's easy to fix, but I wonder why the critical depth limit seems
to be so much less on your machine. I get the expected error up to
about 65000 UNION ALLs --- why is yours crashing at a tenth of that?

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tomas Vondra 2012-11-11 17:32:01 Re: too much pgbench init output
Previous Message Matthew Gerber 2012-11-11 16:30:32 Re: Unresolved error 0xC0000409 on Windows Server