Actually applying workaround for Q323602 did not fixed the problem
completely. The problem was the value for SQLBUFFERSIze - I was too greedy
for performance, set it to maximum (4096) and stability suffered. It
worked but after one day of backups started failing from time to time.
Reducing the value to the default 1024kB resolved the problem and
sacrificed the performance. With sqlbuffsize 4096 and autoset number of
sqlbuffers the rate of backups was 25 MB/s (1 stripe, 2 stripes got 32
MB/s). After changing the threading model the log backup failures
disappeared but the biggest databases were still failing on
full/differential backups. Returning sqlbuffersize back to (default)
1024kB prevented errors from occuring but the speed of backups dropped to
8-8.5 MB/s !?!
The SQL server system is with 8 processors and nearly a GB of RAM free
(from 8 GB). I guess SQL server under load (and this box is really loaded)
does refuse some buffer allocations. Does TDP retry allocation requests or
not I do not know. What I know is - it gave up and did no backups

Zlatko Krastev
IT Consultant






Maria Ilieva <[EMAIL PROTECTED]>
Sent by: "ADSM: Dist Stor Manager" <[EMAIL PROTECTED]>
23.01.2003 14:10
Please respond to "ADSM: Dist Stor Manager"


        To:     [EMAIL PROTECTED]
        cc:
        Subject:        TDP for SQL errors SOLVED


Thanks Del,

Correcting registry entries according to the Microsoft Knowledge Base
article Q323602 help us.

Maria Ilieva

----- Original Message -----
From: "Del Hoobler" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, January 22, 2003 3:30 PM
Subject: Re: TDP for SQL errors


> Maria, (Matt),
>
> This may be a known Microsoft bug, BUG #: 357231 (SHILOH_BUGS).
> Take a look at Microsoft Knowledge Base article Q323602.
>
>   http://support.microsoft.com/default.aspx?scid=kb;en-us;323602
>
> Microsoft documents the workaround for this issue in the
> knowledge base article.
>
> Thanks,
>
> Del
>
> ----------------------------------------------------
>
> Del Hoobler
> IBM Corporation
> [EMAIL PROTECTED]
>
> - Never cut what can be untied.
> - Commit yourself to constant improvement.
>
> =======================================================================
> > .
> > 01/22/2003 09:00:38 ACO5422E Received the following from the MS SQL
> server:
> > 01/22/2003 09:00:38 [Microsoft][ODBC SQL Server Driver][SQL
Server]BACKUP
> LOG is terminating abnormally.
> > Microsoft SQL-DMO (ODBC SQLState: 42000) (HRESULT:0x80040bc5)
> > 01/22/2003 09:00:39 ACO5436E A failure occurred on stripe number (0),
rc
> = 428
> > 01/22/2003 09:00:39 ACO5407E The SQL server aborted the operation.
> > 01/22/2003 09:00:39 Backup of Virgo failed.
> > 01/22/2003 09:00:39 ACO5422E Received the following from the MS SQL
> server:
> > 01/22/2003 09:00:39 [Microsoft][ODBC SQL Server Driver][SQL
Server]BACKUP
> LOG is terminating abnormally.
> > Microsoft SQL-DMO (ODBC SQLState: 42000) (HRESULT:0x80040bc5)
> > 01/22/2003 09:00:40 ACO5436E A failure occurred on stripe number (0),
rc
> = 428
> > 01/22/2003 09:00:40 ACO5407E The SQL server aborted the operation.
> > 01/22/2003 09:00:41 Backup of WaitImport failed.
> > .
>

Reply via email to