Hi Stumpf
I must admit that seeing as TSM does manage to complete the rest of the
backup, It should also retry the failed file. Free space could then be
available.
Did you actually change any of your settings to solve the problem?
Thanks for your reply
Farren :)
"Stumpf, Joachim" <joachim.stumpf AT DATEV DOT DE>@VM.MARIST.EDU> on 02/04/2002
12:17:28
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: Client Message - size estimate exceeded
Hi Farren,
2 weeks ago I had a pmr open at IBM about this problem.
This happens if the storagepool backuppool runs full during your clients
backing up and you have compressalways yes in your dsm.opt.
I opened the pmr at IBM because I was not sure if the file that the TSM
tried to backup was backuped correct.
And the file was not backed up. The TSM also did not do any retries for
this file.
IBM told me that is "works as designed".
What do you think?
Should the TSM retry the file?
Because after the error message the rest of the backup went fine.
Heres the pmr:
Abstract:
question - backuppool full
.
Operating System: OS/390
Product Group: ADSTAR Distributed Storage Manager
.
Environment:
TSM Server 4.1.5 on OS/390 R2.10
.
Problem:
Hi,
I have a question about the error message ANS1311E.
We got this message as the backuppool went full during a backup.
(a Migration was running at this time...)
schedlog looks like:
03/13/2002 22:26:25 Normal File--> 96,649,216 \\file32
\d$\Home\T03381A\archiv.pst [Sent]
03/13/2002 22:26:52 Normal File--> 28,409,856 \\file32
\d$\Home\T03381A\BL.pst ** Unsuccessful **
03/13/2002 22:26:52 ANS1228E Sending of object '\\file32
\d$\Home\T03381A\BL.pst' failed
03/13/2002 22:26:52 ANS1311E Server out of data storage space
03/13/2002 22:26:52 ANS1898I ***** Processed 89,500 files *****
03/13/2002 22:27:36 Normal File--> 181,224,448 \\file32
\d$\Home\T03468A\TG.pst ** Unsuccessful **
03/13/2002 22:27:36 ANS1114I Waiting for mount of offline media.
03/13/2002 22:29:47 Retry # 1 Normal File--> 181,224,448
\\file32\d$\Home\T03468A\TG.pst [Sent]
server log:
22.26.47 STC00778 ANR0534W Transaction failed for session 17 for
node FILE32 (WinNT) - size
22.26.47 STC00778 estimate exceeded and server is unable to obtain
additional space in storage
22.26.47 STC00778 pool
BACKUPPOOL.
Question:
Was the file '\\file32\d$\Home\T03381A\BL.pst' backed up right?
If not...why the TSM doesent retry this file to backup in the
storagepool?
The rest of the backup seemed to be ok.
regards,
Joachim Stumpf
-ORTENREITER, M. MVS-5698TSMVS -L25A/CFETSM-P3S3-02/03/15-05:15--CR
S5> SERVICE GIVEN= 99 SG/99/
S5> SERVICE GIVEN= 99 SG/99/
Asked for the schedulelog, errorlog and any msg's in the activitylog.
Asked also for output from a 'q backup'.
-ORTENREITER, M. MVS-5698TSMVS -L25A/CFWTSM-P3S3-02/03/18-06:44--CR
S5> SERVICE GIVEN= 39 SG/39/
S5> SERVICE GIVEN= 39 SG/39/
D: There is only one object listed in the log, which was not backed up
due to msg ANS1311E Server out of data storage space.
The same time msg ANR0534W Transaction failed for session 17 for node
FILE32 (WinNT) - size estimate exceeded and server is unable to
obtain additional space in storage pool BACKUPPOOL.
Customer confirmed that the backup-pool was filled up to 99 pct. that
time, but later to this msg other objects were backed up as expected
although those were larger in size. For those file there is only
logged: ANS1114I Waiting for mount of offline media.
E: The maxsizethreshold is 250 MB.
The failing object was just 28 MB in size, while the next object was
181 MB in size and it was successfully bu'ed (after mounting a tape).
The backup-pool is 39 GB in size and migr. threshold was set to 60/40
and one migration process could run at a time. (Values are currently
changed to 40/20 and 2 migr. proc. - since that the pbl. did not
reoccure.)
There are 10 nodes scheduled for backup into this pool. All together
those are backing up around 107 GB per night.
A: Suggested to split the schedule or to increase the size of the
backup-pool or to inrease the amount of mountpoints.
Customer should avoid the situation that the stgpool will reach a
utilization of above 99 pct.
..
Found msg ANR0534W s most commonly caused by specifying COMPRESS=YES
and COMPRESSALWAYS=YES in the client dsm.opt file.
Checked customers optionfiles => both are set to YES.
Told customer:
By asking the TSM client to always compress a file, file that have
already been compressed can result in the size of the file increasing
and the backup of the file failing with the messages observed by the
customer.
N: Customer will set COMPRESSALWAYS to NO and req. FUP for tomorrow.
-ORTENREITER, M. MVS-5698TSMVS -L25A/CFBTSM-P3S3-02/03/19-05:59--CC
S5> SERVICE GIVEN= 99 SG/99/
S5> SERVICE GIVEN= 99 SG/99/
The above action solved the pbl, but Mr. Stumpf still comlained that the
client should have to retry the backup of that file.
Found in the msg-guide for ANS1311E System action:
'TSM ended the current operation.'
Mr. Stumpf agreed WaD.
regards / mit freundlichen Gruessen
Joachim Stumpf
Datev eG
Nuremberg - Germany
Farren Minns <fminns AT WILEY.CO DOT UK> (Dienstag, 2. April 2002, 11:15:51,
MEST):
> Hi TSMers
>
> TSM server 3.7.3.8 running on Solaris 2.7 machine. Getting the following
> message for one of our NT client nodes.
>
>
> 04/01/02 22:14:20 ANR0534W Transaction failed for session 43012 for
node
> PUBEARLY (WinNT) - size estimate exceeded
and server is
> unable to obtain additional space in
storage pool BACKUPPOOL.
>
>
> I have looked at the help for the message and from what I can see there
are a number of possible reasons. But I have never seen it
> before for any client and was wondering is anyone out there had had
dealings with this and what remedies they used to solve it. The
> backup eventually completes OK with just one failed file. Unfortunately,
I cannot see which one as the person in charge of the NT machines
> is away on holiday at the moment. I would like to gather some more inf.
before I next speak to him.
>
> Can anyone help with this one.
>
> Many thanks in advance
>
> All the best
>
> Farren Minns - Trainee Solaris & TSM system admin.
> John Wiley & Sons Ltd
>
|
