Re: Out of memory error in 8.1.0 Win32 - Mailing list pgsql-general

From Todd A. Cook
Subject Re: Out of memory error in 8.1.0 Win32
Date
Msg-id 449AD5BC.3060400@blackducksoftware.com
Whole thread Raw
In response to Re: Out of memory error in 8.1.0 Win32  ("Relyea, Mike" <Mike.Relyea@xerox.com>)
Responses Re: Out of memory error in 8.1.0 Win32
List pgsql-general
Hi,

I am consistently running into out-of-memory issues in 8.1.4 running on
RHEL3 and 8.0.5 on RHEL4.  The logs show entries like this:

AggContext: -2130714624 total in 271 blocks; 9688 free (269 chunks); -2130724312 used
TupleHashTable: 893902872 total in 119 blocks; 1088688 free (449 chunks); 892814184 used

which looks mighty suspicious to me. :-;  I can provide a self-contained
test case if anyone wants to look at it.

-- todd


Relyea, Mike wrote:
> So what's my next step?  How do I track down what is causing this
> problem?
>
> -----Original Message-----
> From: Qingqing Zhou [mailto:zhouqq@cs.toronto.edu]
> Sent: Wednesday, June 21, 2006 11:01 PM
> To: Relyea, Mike
> Cc: pgsql-general@postgresql.org; Tom Lane
> Subject: RE: [GENERAL] Out of memory error in 8.1.0 Win32
>
>
>
> On Wed, 21 Jun 2006, Relyea, Mike wrote:
>
>> ExecutorState: 550339936 total in 123 blocks; 195003544 free (740135
>> chunks); 355336392 used
>> HashBatchContext: 293593176 total in 44 blocks; 3107384 free (80
>> chunks); 290485792 used
>> TIDBitmap: 2088960 total in 8 blocks; 924720 free (27 chunks); 1164240
>> used
>> HashBatchContext: 41877664 total in 11 blocks; 4845648 free (15
> chunks);
>> 37032016 used
>> TIDBitmap: 2088960 total in 8 blocks; 1012120 free (27 chunks);
> 1076840
>> used
>
> The same problem. ExecutorState uses much more memory than we expect --
> but not sure where they are from :-(
>
> Regards,
> Qingqing
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: In versions below 8.0, the planner will ignore your desire to
>        choose an index scan if your joining column's datatypes do not
>        match
>


pgsql-general by date:

Previous
From: Tom Lane
Date:
Subject: Re: htonl necesssary for PQexecParams input?
Next
From: Dylan Hansen
Date:
Subject: auto-vacuum & Negative "anl" Values