Hi,

To prevent logical mistake I want to implement standby database which lags behind primary database for certain time. I mean primary db is in operation and copies logs to standby database, while standby database waits certain time (for example 24h) and applies these logs - so basically standby database is 24 hours behind primary db.
The goal is to prevent logical mistake like accidental table deletion which basically doesn't break things but further work is not possible. There are other mechanisms on this system to prevent such errors as node down.

How do I implement that ? I was reading about db2 HADR and there are similarities - but one problem - standby database applies logs in no time - it doesn't wait any period of time. So how do I implement that - by using HADR options, or by other method ?


thanks
Vilius

try log shipping
copy the archived logs to new system and rollforward at any time
to get the db active, you need to bring it out of rollforward state
to continue the log shipping - restore is needed
also with hadr - the standby db can not be accessed while in rollforward mode
maybe replication can help for you and start apply at certain times, so db is always available and up-to-date to whatever moment needed