Hi,

We use db2 8.2.6 on Windows.

Windows clustering along with SAN Discs constitute our highavailability strategy. But recently we had a faiure in SAN that resulted in DB2 unavailability. The management is now requesting a HADR that includes provison for SAN failures. In other words by some magic, they require a stand by database (of course not using SAN disc) to take over when SAN related failure occur.

Have anyone got experiences to share on these type of scenarios? is there any DB2 documentation available that provides some insight on HADR to cover SAN failures.


Thanks

Hari Kumar

For HADR there is no shared disk. Both the primary and standby database each have their own disk mounts. If both of these are on the same SAN, then of course HADR may not help if you have a complete or even partial SAN failure. But if they are on different SAN's you will should be protected if one SAN goes down.

Make sure that the DB2 path names are identical for both the primary and standby database. You could use internal (non-SAN) disk for the standby, but the path names need to be same on both servers for the DB2 database. This is so that DDL commands that reference container names (with an associated path name) will replicate properly on the standby.

Will try this. thanks for suggestion

Hari Kumar