Hi Tilo,
I have found the problem.
We have available 2 versions of backint for Oracle.
Version 5.1.0.135 and 5.1.0.150
Version 5.1.0.135 will work with NetBackup 5.1 MP3.
Version 5.1.0.150 will work with NetBackup 5.1 MP4.
If you try to use 5.1.0.150 with NBU 5.1 MP3, the Backint
for Oracle and the Backint for MaxDB do not communicate properly,
and backup never succeeds.
These versions of backint for Oracle will both work for backing
up an Oracle database with either patch level of NBU, but not
with backint for MaxDB in the picture.
I've not turned up the tracing level on the client, so I don't know
if it would have turned up anything useful.
The symptom was the that backint for MaxDB would continue
to wait on backint for Oracle as seen in dbm.epb.
eg.
Waiting 8 seconds ... Done.
Checking backup tool.
The backup tool is running.
Waiting 9 seconds ... Done.
Checking backup tool.
The backup tool is running.
Waiting 10 seconds ... Done.
Checking backup tool.
The backup tool is running.
...
Incidentally, the dbm.ebp file is not much use in troubleshooting unless
you can catch it on the first backup attempt. NBU tries 3 times before
completely failing a job.
As the first attempt leaves some log files in place, the second attempt
fails because the files already exist, masking the real problem.
The logs are fortunately accumulated in dbm.prt.
Jared
On 12/20/05, Heinrich, Tilo <[EMAIL PROTECTED]> wrote:
>
> Hello Jared,
>
> Could you please send a dbm.ebp from a backup that gets canceled to this
> list? Is there a difference in database sizes?
>
> Best regards,
> Tilo Heinrich
> SAP Labs Berlin
>
> ________________________________
>
> From: Jared Still [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, December 20, 2005 7:58 PM
> To: Heinrich, Tilo
> Cc: [email protected]
> Subject: Re: MaxDB backups
>
>
> Thank you for the reply Tilo.
>
> Yes, it is being killed.
>
> I changed the timeout to 30 seconds so it wouldn't take
> so long for the job to fail. It was taking several minutes
> prevoiusly with the same results, which is why I set the
> timeout value.
>
> The job was working previously on our test NBU server.
>
> The only difference I can find between the 2 is that the
> test server is Veritas 5.1 MP4, while production is 5.1 MP3.
>
> The error and output files get created, the staging file is
> created, but it seems to hang on the communication between
> backint for MaxDB and backint for Oracle.
>
> Jared
>
>
>
> On 12/20/05, Heinrich, Tilo <[EMAIL PROTECTED]> wrote:
>
> Hello Jared,
>
> Please have a look into one of the files dbm.ebp and dbm.ebl.
> Most
> probably you will find, that your Backint got killed by the DBM
> Server
> after a certain timeout elapsed. If so, change the timeout with
> the help
> of parameter "TIMEOUT_SUCCESS" in the configuration file for
> Backint for
> MaxDB (see also
>
> http://dev.mysql.com/doc/maxdb/en/55/a81d36451f11d5992400508b6b8b11/fram
>
> eset.htm).
>
> Best regards,
> Tilo Heinrich
> SAP Lab Berlin
>
>
> -----Original Message-----
> From: Jared Still [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, December 20, 2005 12:38 AM
> To: [email protected]
> Subject: MaxDB backups
>
> Hi All,
>
> I'm currently seeing an issue with backing up a MaxDB database.
>
> This is on Windows Server 2003, MaxDB 7.5, and Veritas
> NetBackup.
>
> While I had this setup on a test server, all worked fine.
>
> Moved the backups to the production NBU server, and now I am
> seeing the following errors in the dbm.prt file
>
> 2005-12-19 13:06:31 0x00000ec8 0 DBM command
> param_startsession
> 2005-12-19 13:06:32 0x00000ec8 0 DBM command
> param_abortsession
> 2005-12-19 13:06:33 0x00000ec8 0 DBM command
> param_startsession
> 2005-12-19 13:06:33 0x00000ec8 0 DBM command
> param_abortsession
> 2005-12-19 13:06:36 0x00000ec8 0 DBM command
> db_execute
> SELECT ID,CREATEDATE,USEDSIZE
> 2005-12-19 13:06:36 0x00000ec8 ERR -24988 DBM ERR_SQL:
> sql
> error
> 0x00000ec8 ERR -24988 DBM 100,Row
> not found
> 2005-12-19 15:04:42 0x00000de0 0 DBM command
> backup_start
> BACKBOOFULL DATA
> 2005-12-19 15:05:58 0x00000de0 ERR -24919 DBM
> ERR_BACKUPOP:
> backup
> operation was unsuccessful
> 0x00000de0 ERR -24919 DBM The
> backup tool
> was
> killed with 0 as sum of exit codes. The database request ended
> with code
> 0.
>
> I've googled for this with a number of different keywords: no
> help
> found.
>
> Any ideas on what the basic problem is here?
>
> Thanks,
>
> --
> Jared Still
> Certifiable Oracle DBA and Part Time Perl Evangelist
>
> --
> MaxDB Discussion Mailing List
> For list archives: http://lists.mysql.com/maxdb
> To unsubscribe:
> http://lists.mysql.com/[EMAIL PROTECTED]
>
>
>
>
>
>
> --
> Jared Still
> Certifiable Oracle DBA and Part Time Perl Evangelist
>
>
>
--
Jared Still
Certifiable Oracle DBA and Part Time Perl Evangelist
