rees_chris
ADSM.ORG Member
Hi
We've got a problem with a particular DB2 backup that I'm hoping someone can help with.
TSM Server 5.5.1 (AIX)
DB2 Server is AIX. TSM Client 5.5.1
DB2 version 9.1
At night we do a full db2 backup including logs. This is run LAN-Free to LTO3 tape.
I'll use last night as an example,
What we see is the following:
db2 backup starts
backup got to 336Gb complete
Then , proxy session goes RecvW and STA session goes IdleW.
The sessions then sit there until resourcetimeout is hit which is default 60.
So after 60 minutes I see "ANR0538I A resource waiter has been aborted." in the tsm actlog
Hence there are no tsm sessions left, but in DB2 it thinks the backup is still running as shown by db2 list applications show detail.
Its almost like DB2 just stops sending data. In the db2diag.log I can see no errors.. there are lots of messages similar to that below but then they just stop at the same time that the sessions went IdleW/RecvW
2008-08-29-12.43.18.605506+000 I3190272873A441 LEVEL: Info
PID : 1503268 TID : 1 PROC : db2bm.2224166.15 0
INSTANCE: db2pmq NODE : 000
APPHDL : 0-2054 APPID: *LOCAL.db2pmq.080829120641
AUTHID : SAPPMQ
FUNCTION: DB2 UDB, database utilities, sqluWriteToQueue, probe:333
MESSAGE : Sending message
DATA #1 : String, 14 bytes
SqluBufAddrMsg
DATA #2 : Hex integer, 4 bytes
0x02D000A2
The strange thing is , if we kill the session in DB2 and then run it manually during the day it seems to work fine.
I'm guessing there is either a lock in the DB2 database or the TSM database that is not released.
I've got 15 other DB2 backups at the same versions backing up in the same way that have never had this problem. Its always just the one database that does this.
Anyone got any ideas.?
cheers
We've got a problem with a particular DB2 backup that I'm hoping someone can help with.
TSM Server 5.5.1 (AIX)
DB2 Server is AIX. TSM Client 5.5.1
DB2 version 9.1
At night we do a full db2 backup including logs. This is run LAN-Free to LTO3 tape.
I'll use last night as an example,
What we see is the following:
db2 backup starts
backup got to 336Gb complete
Then , proxy session goes RecvW and STA session goes IdleW.
The sessions then sit there until resourcetimeout is hit which is default 60.
So after 60 minutes I see "ANR0538I A resource waiter has been aborted." in the tsm actlog
Hence there are no tsm sessions left, but in DB2 it thinks the backup is still running as shown by db2 list applications show detail.
Its almost like DB2 just stops sending data. In the db2diag.log I can see no errors.. there are lots of messages similar to that below but then they just stop at the same time that the sessions went IdleW/RecvW
2008-08-29-12.43.18.605506+000 I3190272873A441 LEVEL: Info
PID : 1503268 TID : 1 PROC : db2bm.2224166.15 0
INSTANCE: db2pmq NODE : 000
APPHDL : 0-2054 APPID: *LOCAL.db2pmq.080829120641
AUTHID : SAPPMQ
FUNCTION: DB2 UDB, database utilities, sqluWriteToQueue, probe:333
MESSAGE : Sending message
DATA #1 : String, 14 bytes
SqluBufAddrMsg
DATA #2 : Hex integer, 4 bytes
0x02D000A2
The strange thing is , if we kill the session in DB2 and then run it manually during the day it seems to work fine.
I'm guessing there is either a lock in the DB2 database or the TSM database that is not released.
I've got 15 other DB2 backups at the same versions backing up in the same way that have never had this problem. Its always just the one database that does this.
Anyone got any ideas.?
cheers
