Hi,
I'm just transferring a database to V8.1.5 and considering using a standby
database on a remote host to provide a hot backup in case I lose my
production database. The remote will be on the same LAN so transmission
times should not be a problem.
I've read through all the docs on the subject and am left with a couple of
questions which I can't seem to get an answer to.
- Is there a 'recommended' way to calculate the size that redo logs should
be? I have tended to accept the default size of around 500k and have three
of them.
- I do a full cold database backup once per week, at which time I move the
archive directory to one side and create a new empty directory. I then
backup the database files and the 'old' redo logs to tape and then delete
the old redo log directory. Is there any reason for keeping this directory
around any longer? The docs are a bit short on advice about how long to hang
on to archived redo logs.
Do the archived logs received by the standby database end up in the
log_archive_dest for that database? What do I do about clearing them away if
so?
- It seems that I can set an archive destination to my remote standby
database as OPTIONAL. Why would I want to do that? The documentation seems a
little unclear about this. I want the primary database to hang up if the
standby cannot be updated (presumably giving CANNOT CONNECT UNTIL FREED type
messages) until all the archived logs have been transferred. The doc seems
to suggest that the primary database would give up trying if the remote
database is unavailable and would overwrite my redo logs. Does setting the
destination to MANDATORY mean that the primary will hang until it has been
able to transmit the logs to my standby?
Any comments on the above would be most appreciated! I am going to give this
a good test out but would appreciate a better understanding of what is going
on!!
Cheers,
Graham Potts
Received on Tue Feb 08 2000 - 16:04:55 CST
