TSM 6.1 Instance & DB Backup

[paschumacher]

I currently have a TSM server running Red Hat ES 5 with TSM6.1.1 and I just created a 2nd instance of TSM on the same physical server in a shared library scenario, both have the same IP addresses, but different port numbers.

I am trying to manually backup the DB on the new TSM server and getting this in act log:

08/20/2009 14:36:12 ANR2017I Administrator PASCHUMACHER issued command: BACKUP
DB dev=3584class type=full (SESSION: 12)
08/20/2009 14:36:12 ANR4559I Backup DB is in progress. (SESSION: 12)
08/20/2009 14:36:12 ANR0984I Process 43 for DATABASE BACKUP started in the
BACKGROUND at 02:36:12 PM. (SESSION: 12, PROCESS: 43)
08/20/2009 14:36:12 ANR2280I Full database backup started as process 43.
(SESSION: 12, PROCESS: 43)
08/20/2009 14:36:14 ANR4578E Database backup/restore terminated - required
volume was not mounted. (SESSION: 15)
08/20/2009 14:36:15 ANR2980E Database backup terminated TCPIP failure - DB2
sqlcode -2033 sqlerrmc -50. (SESSION: 12, PROCESS: 43)
08/20/2009 14:36:15 ANR0985I Process 43 for DATABASE BACKUP running in the
BACKGROUND completed with completion state FAILURE at
02:36:15 PM. (SESSION: 12, PROCESS: 43)

any help would be appreciated. thanks!

[chad_small]

Are those the only errors? Do you have scratch available? Is it possibly trying to mount a tape and the library inventory is screwed up, or does this only happen when you run the DB backup?

[paschumacher]

yes those are the only errors (so far) i have 56 scratch tapes and so far this has only happened when I try to do a backup of the library client. i havent tried to backup anything else yet.

[picay]

What is the configuration of device class 3584class ?
Can you post output of q devc 3584class f=d please ?

[paschumacher]

From the Library Manager:
tsm: CRONUS>q dev 3584class f=d

Device Class Name: 3584CLASS
Device Access Strategy: Sequential
Storage Pool Count: 0
Device Type: LTO
Format: DRIVE
Est/Max Capacity (MB):
Mount Limit: DRIVES
Mount Wait (min): 60
Mount Retention (min): 2
Label Prefix: ADSM
Library: 3584LIB
Directory:
Server Name:
Retry Period:
Retry Interval:
Shared:
High-level Address:
Minimum Capacity:
WORM: No
Drive Encryption: Allow
Scaled Capacity:
Last Update by (administrator): PASCHUMACHER
Last Update Date/Time: 06/01/2009 16:25:53

From the Library Client
tsm: TSMPS>q dev 3584class f=d

Device Class Name: 3584CLASS
Device Access Strategy: Sequential
Storage Pool Count: 2
Device Type: LTO
Format: DRIVE
Est/Max Capacity (MB):
Mount Limit: DRIVES
Mount Wait (min): 60
Mount Retention (min): 60
Label Prefix: ADSM
Library: 3584LIB
Directory:
Server Name:
Retry Period:
Retry Interval:
Shared:
High-level Address:
Minimum Capacity:
WORM: No
Drive Encryption: Allow
Scaled Capacity:
Last Update by (administrator): PASCHUMACHER
Last Update Date/Time: 08/20/2009 12:27:58

[picay]

and output of q libr f=d on library client please ?
from library client, did you "ping server" the primary library manager ?
I do mind about the tcpip failure:

08/20/2009 14:36:15 ANR2980E Database backup terminated TCPIP failure - DB2
sqlcode -2033 sqlerrmc -50. (SESSION: 12, PROCESS: 43)

[paschumacher]

yes i can ping server and communicate back and forth.

tsm: TSMPS>ping server cronus
ANR1706I Ping for server 'CRONUS' was able to establish a connection.

tsm: CRONUS>ping server tsmps
ANR1706I Ping for server 'TSMPS' was able to establish a connection.

output of q libr from TSMPS (library client)

tsm: TSMPS>q libr f=d

Library Name: 3584LIB
Library Type: SHARED
ACS Id:
Private Category:
Scratch Category:
WORM Scratch Category:
External Manager:
Shared: No
LanFree:
ObeyMountRetention:
Primary Library Manager: CRONUS
WWN:
Serial Number:
AutoLabel:
Reset Drives:
Relabel Scratch:
Last Update by (administrator): PASCHUMACHER
Last Update Date/Time: 08/20/2009 15:59:26

[picay]

yep, message meaning is clear, tcpip problem, so...
It happend only for database backup ?
Since the begining of installation of library sharing ?
or did dbbackup runs fine before with the same configuration and problem appear recently ?
Between the last good backup and the first erroneous one, what happend on network configuration ?
Maybe I take a wrong way but... I do mind about message ANR2980E

[paschumacher]

IBM doesn't seem to have an answer either. This is a new TSM 6.1 instance on the same physical server as the library manager, using the same IP but different TCP port #. I have only tried a DB backup so far. I don't have any nodes configured yet. I have always tested my lib client to lib manager comms with a manual DB backup before i kick anything else off.

[kehannigan]

The error is from DB2, calling the TSM API to back up to the server. Return code -50 means that the API couldn't connect to the server. So the first thing to check is that the API is configured correctly, and that DB2, in turn, is configured to use the API correctly.

There's a topic in the V6 library that goes through the necessary steps to do this. Hopefully it'll help pinpoint just what's going wrong...

http://publib.boulder.ibm.com/infoce...rep_dbmgr.html

[john0881]

This is just a guess...but have you double-checked the DSMI_CONFIG environment variable that is set in the ".../sqllib/userprofile"? If you don't have it set, make sure you set it and recycle the TSM instance and try it again.

Like I said...only a guess.

[paschumacher]

actually got an update from IBM on this issue. it looks like I defined the tape drives incorrectly being this is now an instance of TSM rather than a stand-alone "server" I defined the drives the old school way.

I asked why it was saying TCP/IP communication failure for something like tape drives, and they couldn't tell me.

[issen]

Hi,
Here is a solution that solve my DB2 problem

http://publib.boulder.ibm.com/infoce...rep_dbmgr.html

[unpaired]

I had smilar problem. Problem is multipath problem. Delete fsci path and than define again it. Problem is solved.