Hello,
That makes perfect sense, if I say LIST BACKUP you'd expect the controlfile
and catalog to be brought into line before any results are returned.
I also found the rman options "DEBUG TRACE=/tmp/1.txt" handy when trying to
understand whats going on but are very verbose, have plenty of disk space.
Hopefully you just need to change the query to pick the "last" SCN from
your existing backup but I remember having to go up and down from that SCN
to get the restore to work.
Regards,
Frank
On Fri, Aug 7, 2020 at 4:25 PM Charlotte Hammond <
charlottejanehammond@xxxxxxxxx> wrote:
Thanks Frank.
Just connecting doesn't seem to be enough to cause a resync but most other
things do.
We have decided to bite the bullet and attempt to address this problem
with our duplicate utility - it's just too flakey to cross our fingers and
hope nobody accidentally does an unintentional resync just before our
duplicate is due to start.
Thank You!
Charlotte
On Thursday, August 6, 2020, 11:11:29 AM GMT+1, Frank Gordon <
frankagordon@xxxxxxxxx> wrote:
Hello,
can you equate anything like archivelog backup or your browsing sessions
to the timestamps for
"periodically updated over the course of the day"
I'd also try running SQL*PLUS queries on the catalog rather than RMAN list
commands,
that'd cut out one source of implicit resync.
Regards,
Frank
On Mon, Aug 3, 2020 at 8:22 PM Charlotte Hammond <
dmarc-noreply@xxxxxxxxxxxxx> wrote:
Hello!
What causes an implicit resync of the RMAN catalog? I had thought it was
significant events like running backups or deleting backup pieces. However
I've been using RMAN today, connected to the catalog, just browsing but I
can see that the catalog has periodically updated over the course of the
day.
This shouldn't be a problem of course, but we have a legacy utility that
runs a duplicate from backups without specifying the SCN. In this case the
SCN seems to be plucked from the highest SCN in the AL table. This is
fine if the AL table isn't updated outside of backups but otherwise it ends
up with SCNs from more recent archive logs which aren't in the backup and
so the duplicate fails.
Any workarounds without hacking at the fragile old duplicate utility to
force it to use a specific SCN or time?
Thank you in advance for any advice! (Oracle 11.2)
Charlotte
--
+353-86-0695383
