DB2 archive log and online backup targets

jamesmacd40

ADSM.ORG Member
Joined
Apr 17, 2018
Messages
67
Reaction score
3
Points
0
PREDATAR Control23

Hi all

i am trying to figure out how to handle DB2 databases/log backups in a new tsm setup (not tsm DB). The current set up is that on the prod side, the DB2 archive logs are being dumped to a random access storage pool in TSM (as are full online DB backups in its own stgpool). from there it is moved to a tapepool and then a copy tape pool.
The environment uses HADR, so it replicates DB2 log updates over the WAN to the DR site. DB2 is designed during this HADR failover/failback to reference a single TSM server instance(the prod server- which is restored to the DR server) which allows for automatic backup and recovery without having to change anything in the HADR setup

If i use TSM replication, which i want to do for the new setup, i do not know how this will affect the DB2 processes for DR. obviously the DR tsm server will have its own name/instance so the design reference to a single TSM server instance will not work. I am not sure what type of stgpool to use for logs/db - do i/should i define separate container storage pools for these and replicate them or some other method?

i am a bit lost on this design aspect.
 
PREDATAR Control23

You should be ok because the baclient and the API can automatically failover to the DR server for restore purpose only:
https://www.ibm.com/support/knowledgecenter/SSEQVQ_8.1.4/client/c_autoclientfailover.html
https://www-01.ibm.com/support/docview.wss?uid=swg21649484

You won't be able to continue to archive the logs though on the DR side.

Technically, you can if you change the roles of DR server to become the primary server. And then you flip the roles back when you bring the other server online. It's a lot of work, so you would not do this for a short outage, you would only do this if the primary server is destroyed and will take several days to rebuild it:
http://www-01.ibm.com/support/docview.wss?uid=swg21628618
 
PREDATAR Control23

That may be a problem if DR happens and the logs need to write to TSM during that period, which no doubt they will.
So this begs to question what type of landing storage pool do i use for DB2 archive logs and the full online DB backups. currently it is just a diskpool that fills to 80% then migrates to tape.
wondering if i can still do a diskpool that will migrate to container storage pool that replicates
 
PREDATAR Control23

That may be a problem if DR happens and the logs need to write to TSM during that period, which no doubt they will.
So this begs to question what type of landing storage pool do i use for DB2 archive logs and the full online DB backups. currently it is just a diskpool that fills to 80% then migrates to tape.
wondering if i can still do a diskpool that will migrate to container storage pool that replicates
Same issue, you can't do backup to node on a server when that server is a target server.
 
Top