ADSM-L

Re: ADSM-L Digest - 8 Nov 2001 to 9 Nov 2001 (#2001-306)

2001-11-10 09:06:21
Subject: Re: ADSM-L Digest - 8 Nov 2001 to 9 Nov 2001 (#2001-306)
From: Ken Chamberlain <ken.chamberlain AT UTORONTO DOT CA>
Date: Sat, 10 Nov 2001 09:03:27 -0500
While I don't run multiple TDP-R3's. I thought I might make some
observations:
You are running the TDP-R3 backups from crontab you say, so there are three
crontab entries, one for each ora<sid> that look like "brbackup -<some
parameters>" or at least call a script which executes brbackup.  The current
directory is ora<sid>'s home directory unless changed by a script that calls
brbackup.  You should be looking there for ds*error.log.  Note that backint
uses the TSM API, thus expect dsierror.log not dsmerror.log.  Also check
dsm.sys and dsm.opt in /usr/tivoli/tsm/client/api/bin although they may link
to ../../ba/bin/.  I am not sure why you code domain statements for your sap
log and data filesystems.  This suggests you are taking duplicate (I might
note unusable) backups of log and data files in your incremental backups,
unless you exclude them, thus meaning you are just backing up the directory
structure in the log and data filesets.

I experienced problems with the TDP-R3 full offline backup at the 3.2.0.3
level that were all traced back to TDP-R3's use of memory.  Try a newer
version.  The only version I have successfully used in production is
3.2.0.4.  I've tried 3.2.0.5 but I have problems at that level.  I hear
3.2.0.6 is out but I have yet to try it.

You must have a LOT of tape drives to run 3 TDP-R3 at the same time.  Try
summing the SESSIONS parameter in each $ORACLE_HOME/dbs/init<SID>.utl file.
If that sum exceeds the number of tape drives you have on TSM server, expect
problems such as you are experiencing.  The solution is to run TDP-R3
serially.

Ken Chamberlain
University of Toronto
Phone:(416) 978-1582
Fax:  (416) 971-2085
E-mail: ken.chamberlain AT utoronto DOT ca

----- Original Message ----- >
> Date:    Fri, 9 Nov 2001 15:18:54 -0600
> Date:    Fri, 9 Nov 2001 15:18:54 -0600
> From:    "Davidson, Becky" <Becky.Davidson AT SLBG DOT COM>
> Subject: Re: TSM/TDP-R3 Question
>
> What happens when you run each separately?
> When you restart them do you restart via sapdba or reschedule the cron?
> Make sure that the profile is correct for the user scheduling the cron
> job....perhaps even running the profile as part of the command line or
> setting all of the important variables in the script that kicks off the
> backup.
> Good Luck
> Becky
>
> -----Original Message-----
> From: Rosa Leung/Toronto/IBM [mailto:rosal AT CA.IBM DOT COM]
> Sent: Friday, November 09, 2001 2:42 PM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: TSM/TDP-R3 Question
>
>
> No, I don't see any BK messages.
>
> DSM.SYS
>
> SErvername  TORTSM03
> *TCPBUFFSIZE 32
> TCPBUFFSIZE  256
> *TCPWINDOWSIZE 64
> TCPWINDOWSIZE 1024
> TXNBYTELIMIT  25600
> TCPNODELAY YES
> PASSWORDACCESS GENERATE
> SCHEDLOGRETENTION 14
> SCHEDLOGNAME "/usr/tivoli/tsm/client/ba/bin/dsmsched.log"
> COMPRESSION  ON
> ERRORLOGRETENTION 14
> ERRORLOGNAME "/usr/tivoli/tsm/client/ba/bin/dsmerror.log"
> INCLEXCL "/usr/tivoli/tsm/client/ba/bin/inclexcl.lst"
>    COMMmethod         TCPip
>    TCPPort            1500
>    TCPServeraddress   10.39.5.19
> *NODENAME TORTSM03
>
> DSM.OPT
>
> SE TORTSM03
> DOMAIN /
> Domain /usr
> Domain /var
> Domain /home
> Domain /performance
> Domain /ibm
> Domain /celestica
> Domain /util
> Domain /oracle/ES1
> Domain /usr/sap/ES1
> Domain /sapmnt/ES1
> Domain /oracle/stage
> Domain /oracle/ES1/origlogA
> Domain /oracle/ES1/origlogB
> Domain /oracle/ES1/mirrlogA
> Domain /oracle/ES1/mirrlogB
> Domain /oracle/ES1/saparch
> Domain /oracle/ES1/sapreorg
> Domain /oracle/ES1/sapdata1
> Domain /oracle/ES1/sapdata2
> Domain /oracle/ES1/sapdata3
> Domain /oracle/ES1/sapdata4
> Domain /oracle/ES1/sapdata5
> Domain /oracle/ES1/sapdata6
> Domain /oracle/ES1/sapdata6
> Domain /oracle/ES1/sapdata6
> Domain /oracle/ED1
> Domain /usr/sap/ED1
> Domain /sapmnt/ED1
> Domain /oracle/ED1/origlogA
> Domain /oracle/ED1/origlogB
> Domain /oracle/ED1/mirrlogA
> Domain /oracle/ED1/mirrlogB
> Domain /oracle/ED1/saparch
> Domain /oracle/ED1/sapreorg
> Domain /oracle/ED1/sapdata1
> Domain /oracle/ED1/sapdata2
> Domain /oracle/ED1/sapdata3
> Domain /oracle/ED1/sapdata4
> Domain /oracle/ED1/sapdata5
> Domain /oracle/ED1/sapdata6
> Domain /oracle/EPS/origlogA
> Domain /oracle/EPS/origlogB
> Domain /oracle/EPS/mirrlogA
> Domain /oracle/EPS/mirrlogB
> Domain /oracle/EPS/saparch
> Domain /oracle/EPS/sapreorg
> Domain /oracle/EPS/sapdata1
> Domain /oracle/EPS/sapdata2
> Domain /oracle/EPS/sapdata3
> Domain /oracle/EPS/sapdata4
> Domain /oracle/EPS/sapdata5
> Domain /oracle/EPS/sapdata6
> subdir yes
>
> Thanks
>
>
>
> Rick Kylie
> 245 Consumers Road  M2J 1R3  ON
> Tel: (416) 490-2121 Fax: (416) 490-5283
> Rkylie AT ca.ibm DOT com
> Loc: BK02/Con/245/TOR
>
>
> "Davidson, Becky" <Becky.Davidson AT SLBG DOT COM>@VM.MARIST.EDU> on 
> 11/09/2001
> 12:04:12 PM
>
> Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
>
> Sent by:  "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
>
>
> To:   ADSM-L AT VM.MARIST DOT EDU
> cc:
> Subject:  Re: TSM/TDP-R3  Question
>
>
> Do you get any BKI messages or any additional messages?  What do your
> dsm.sys and dsm.opt files look like?
>
> -----Original Message-----
> From: Rosa Leung/Toronto/IBM [mailto:rosal AT CA.IBM DOT COM]
> Sent: Friday, November 09, 2001 10:45 AM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: TSM/TDP-R3 Question
>
>
> -We are running TSM and using TDP-R3 for SAP .
>
> Storage Management Server for AIX-RS/6000 - TSM Version 4, Release 1,
Level
> 3.0
>
> Tivoli/TSM.TDP/R3 Version 3, Release 2, Level 0.3  for AIX 4.3.3
>
> We have some servers with more then one SAP  DataBase instance on them and
> we are see the following errors
>
> BR232E 0 of 85 files saved by backup utility
> BR280I Time stamp 2001-11-09 00.30.15
> BR231E Backup utility call failed
>
> Note:  All three backups start at the same time and fail.
>        They are scheduled to run from CRONTAB
>        After the failure they can be started again and run fine.
>        No errors are seen from the TSM server side.
>
> Has anybody seen this happen before.
>
> Thanks
>
>
> Rkylie AT ca.ibm DOT com
> Loc: BK02/Con/245/TOR
>
<Prev in Thread] Current Thread [Next in Thread>
  • Re: ADSM-L Digest - 8 Nov 2001 to 9 Nov 2001 (#2001-306), Ken Chamberlain <=