Re: Dataguard / Archive Logs

It depends.

If using ARCH as the means of transport, yes, you're probably right. I haven't tested it, nor time to do it the coming days.

If using LGWR to get the redo sent to the standby (using standby redo log files), the logswitch will be performed locally on the standby. So, there is no incomplete archived redo logfile at the standby.

A test for the first case can go along the following steps:

- stop the log-transfer (defer the log_archive_dest_x)
- perform a logswitch on the primary.
- Copy the archived logfile to the standby manually
- blank or delete the last blocks, eg using dd. (talking *n*x here)
- register the logfile manually to the stanby
- try to apply it.
- see what happens

The detailed scenario will also determine when the archived logfile gets registered at the standby. Probably/most likely after the transfer has completed succesfully.

If you get a chance to do the test, I'd be glad if you can share the results. Meanwhile, it's on my 'wishlist' for future tests. Thank you for the question.

Regards, Carel-Jan

===
If you think education is expensive, try ignorance. (Derek Bok) ===

> Hi can someone confirm (or deny) my assumption here.
>
> An archived log is currently being sent from the primary to the
> standby. Primary site goes mid transfer, we end up with a 90%
> completed log on the standby side.
>
> It tries to get applied to the standby. Does oracle reject the whole
> file or can it apply the work up to the dodgy part.
>
> I am sure it will reject the whole file but want confirmation.
>
> Thanks
> --
> http://www.freelists.org/webpage/oracle-l
>

--
http://www.freelists.org/webpage/oracle-l