Re: Standby database, archive logs only get shipped once

Noons wrote:
> Joel Garry apparently said,on my timestamp of 23/02/2005 10:51 AM:
>
> >
> > I've seen weirder things happen when the RFS process gets hosed by
a
> > glitch in the network. Support couldn't help me, one time had to
> > resetlogs and recreate the standby. YMMV.
>
> Is there no way to restart the bugger?

That particular time, bouncing the instance would merely move to the next redo. It was one of the more amazing brain-damaged effects I've seen. Hopefully limited to 8i. metalink 255959.1 for some additional details, that don't really explain why this should affect the primary. But of course, they're still finding bugs, like http://metalink.oracle.com/metalink/plsql/ml2_documents.showFrameDocument?p_database_id=BUG&p_id=4067967 (search the bug db for RFS, get lots of hits).

Hmm, here's one where it works the first time but not after: http://metalink.oracle.com/metalink/plsql/ml2_documents.showFrameDocument?p_database_id=BUG&p_id=2754645 , so unsatisfying when there is no resolution.

>
> >
> > After 3 times in 3 months, customer had me write a simple log
shipper.
> >
>
> Darn! That's not supposed to happen.

Hey, when the standby shoots the primary, you do what you must. I'm still amazed whoever wrote the [v8] managed standby didn't design for common network failures. Why anyone writes HA limited to a local environment is beyond me, anyways. You wind up with something more complicated than useful for most sites.

jg

--
@home.com is bogus.
RIP Gene Scott.  You may have been a fruitcake, but at least you took
on the FCC! http://www.rotten.com/library/bio/religion/dr-gene-scott/
http://hosted.ap.org/dynamic/stories/O/OBIT_SCOTT?SITE=NCGRE&SECTION=HOME&TEMPLATE=DEFAULT