Hi,

was wondering if anyone has any tested/proven scripts for monitoring a
managed standby database on AIX? I'm running 8.1.7.4, with the
primary & standby in separate geographic locations. I've developed a
few scripts of my own to monitor various aspects of the standby log
propogation & apply process, but each one seems to lead to another
undiscovered / undetected issue, which if not caught in time forces me
to rebuild the standby. Need scripts to make sure the logs are being
created at the remote location (it's an optional destination so as not
to hang up my primary), to make sure they are getting applied okay,
and to cleanup the logs but only after they've been applied. Any help
is appreciated.

Thanks
Roy

rspeaker@yahoo.com (Roy) wrote in message news:<498db9a0.0405260727.3526bdea@posting.google. com>...
> Hi,
>
> was wondering if anyone has any tested/proven scripts for monitoring a
> managed standby database on AIX? I'm running 8.1.7.4, with the
> primary & standby in separate geographic locations. I've developed a
> few scripts of my own to monitor various aspects of the standby log
> propogation & apply process, but each one seems to lead to another
> undiscovered / undetected issue, which if not caught in time forces me
> to rebuild the standby. Need scripts to make sure the logs are being
> created at the remote location (it's an optional destination so as not
> to hang up my primary), to make sure they are getting applied okay,
> and to cleanup the logs but only after they've been applied. Any help
> is appreciated.
>

After 3 times in 3 months having some network glitch affect the
production db (once requiring a resetlogs to get past!) my customer
decided to have me automate the old manual way. The init may _say_
"service=xxxx optional reopen=60", but it doesn't necessarily mean it,
there's some basic design flaw in the RCS stuff IMO. I wouldn't even
bother with managed until a later 9i, unless you have very
understanding management or an unusually good network.

No matter what, manual expertise is still necessary with standby. You
might want to grep the alert log for the last Log Compoleted

http://metalink.oracle.com/metalink/..._id= 233425.1
for dataguard issues.

jg
--
@home.com is bogus.
Sergeant Shultz! http://www.signonsandiego.com/uniont...b26grasso.html