Networker

Re: [Networker] critical problem .. help me

2005-06-28 06:33:41
Subject: Re: [Networker] critical problem .. help me
From: Allan Nelson <an AT CEH.AC DOT UK>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Tue, 28 Jun 2005 11:32:16 +0100
It's a long time since I tried this so details are hazy, but we had the same 
problem.
We had a script called to close down Oracle, run the backup, then restart 
Oracle using this pre and post command thing.

It seems though that this pre and post thing works 'for each partition'.  It 
may be thats because we named the partitons to backup (rather than use All).
So, lets say you had 3 partitions on your Oracle server to backup...
/data1
/data2
/data3

What happened with us was that the backup run would start, Oracle was closed 
down - OK so far.
/data1 started backing up
Networker sees there's another partition to backup and so tries to close Oracle 
down (again!)
Meanwhile, /data1 has finished and Oracle is started, but it can't because it's 
trying to be closed down again.
Then /data3 starts - and by this time things are in a mess.

As I say, I haven't looked at this for a few years so apologies if it's a red 
herring.  It may be that if you use 'All', it works as you'd expect, but from 
memory I'm sure we tried that.
We did try just naming 1 partiton e.g. /data1 and then everything worked 
perfectly - apart from the fact we didnt have all the data we needed backed up!

I'd also be interested to know if this behaviour has changed and if anyone has 
a neat solution.

Cheeers... Allan.

DISCLAIMER: Any views or opinions expressed in this e-mail or any files 
transmitted with it are those of the author and do not represent the views of 
NERC unless otherwise explicitly stated. The information contained in this 
e-mail may be subject to public disclosure under the Freedom of Information Act 
2000. Unless the information is legally exempt from disclosure, the 
confidentiality of this e-mail and your reply cannot be guaranteed. 

>>> Andrew Watkins <andrew AT DCS.BBK.AC DOT UK> 28/06/05 10:59:44 >>>
I think some people are missing the point!

I beleive that the problem is not to do with tape block size, but why 
is it that the pstcmd (POST COMMAND) has started before the backup 
has completed? 

Are you sure that the scripts are starting before backup is complete, since
it may be just that the networker index is being backed up and not the
oracle datafiles!

(I am also interested in this solution!)

Andrew


> 
> Also, you can set size of block manually by 
> 
> NSR_DEV_BLOCK_SIZE_DLT=64
> export NSR_DEV_BLOCK_SIZE_DLT
> 
> 
> Petr Hanzlík
> COMA ZÁLOHOVACÍ SYSTÉMY a.s.
> Vítkovická 1
> Ostrava 1       702 00
> Czech Republic
> Tel.:  ++420-59-6656377
> Fax.: ++420-59-6656668
> mailto:phanzlik AT coma DOT cz 
> http://www.coma.cz/ 
> 
> 
> 
> 
> 
> 
> snoop <ksdsya AT RHEA-INFO DOT COM>
> Odeslal: Legato NetWorker discussion <NETWORKER AT LISTSERV.TEMPLE DOT EDU>
> 28.06.2005 03:42
> Odpovìzte prosím u¾ivateli Legato NetWorker discussion; Odpovìzte prosím 
> u¾ivateli snoop
>  
>         Komu:   NETWORKER AT LISTSERV.TEMPLE DOT EDU 
>         Kopie: 
>         Pøedmìt:        critical problem .. help me
> 
> 
> Server : compaq5.1 (tru64 cluster Oracle Server)
> Networker 6.1 (build 186) network/10
> Tape Library : compaq (DLT8000) 2drives
> device name : /dev/ntape/tape6_d1
> /dev/ntpae/tape7_d1
> 
> I back up Oracle Cluster datafile. Datafile size is 300Gbytes
> Oracle script is used for oracle backup in savepnpc.
> the script is easy coding by ksh shell
> 
> ora-begin.sh
> ------------
> 
> #!/usr/bin/ksh
> svrmgrl << EOF
> connect internal
> alter database system begin backup ;
> alter database user1  begin bakcup;
> ...
> ...
> ...
> << EOF
> 
> ora-end.sh
> -----------
> 
> #!/usr/bin/ksh
> svrmgrl << EOF
> connect internal
> alter database system end backup ;
> alter database user1  end bakcup;
> ...
> ...
> ...
> << EOF
> 
> 
> Oracle backup group is Oracle
> this group's savepnpc is in /nsr/res/Oracle.res
> this file consist next syntax
> precmd: "/nsr/script/ora-begin.sh"
> pstcmd: "/nsr/script/ora-end.sh"
> timeout "12:00:00"
> 
> Oracle backup started from auto schedule.
> Oracle Backup progressed normally.
> But before backup is end pstcmd is running and ora-begin shell is started
> this happen in 06:00, backup'end time is 08:00
> Why pstcmd is running before backup is end ?
> backup result in sucess but this data can't recover . because
> ora-end.sh is runned before backup is end.
> help me ...
> 
> and i show messages in /nsr/logs/daemons.log, messages
> 
> Block size is 32768 bytes not 98304bytes, verify the device configuration
> tape positioning by record is disabled
> 
> what's mean ?
> 
> --
> Note: To sign off this list, send a "signoff networker" command via email
> to listserv AT listserv.temple DOT edu or visit the list's Web site at
> http://listserv.temple.edu/archives/networker.html where you can
> also view and post messages to the list. Questions regarding this list
> should be sent to stan AT temple DOT edu 
> =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
> 
> --
> Note: To sign off this list, send a "signoff networker" command via email
> to listserv AT listserv.temple DOT edu or visit the list's Web site at
> http://listserv.temple.edu/archives/networker.html where you can
> also view and post messages to the list. Questions regarding this list
> should be sent to stan AT temple DOT edu 
> =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu 
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

<Prev in Thread] Current Thread [Next in Thread>