Thanks for the question, Inger.
Asked: December 01, 2014 - 3:25 pm UTC
Answered by: Tom Kyte - Last updated: December 01, 2014 - 4:40 pm UTC
Category: Database - Version: 11g
Viewed 1000+ times
In the manual I find stated that I must have one more standby redo log file per thread than I have ordinary redo log files, but no reason WHY.... can you explain a little bit?
here is the ref to the manual: http://docs.oracle.com/cd/E11882_01/server.112/e41134/log_transport.htm#SBYDB4751
and here the statement of the fact:
The standby redo log must have at least one more redo log group than the redo log at the redo source database, for each redo thread at the redo source database.
Looking forward to the WHY.... best regards Inger
and we said...
It is a performance/buffering thing - to make sure the standby can keep up with buffering the redo log information at the standby site in periods of high activity on the source database.
Is this answer out of date? If it is, please let us know via a Review
if the source database has 2 groups - #1 and #2 - and is switching frequently during a burst in activity - we want to make sure the standby can keep up. Se we want it to have at least groups #1, #2, #3 so it can switch from #2 to #3 instead of #2 to #1 if it has to. That gives the standby a little breathing room to finish everything it needs to do against group #1 before it can be reused...