Re: 8.1.7 shutdown

"HansF" <News.Hans_at_telus.net> wrote in message news:pan.2005.09.08.14.37.18.72445_at_telus.net

>
> What are the exact steps you are using to shut down the database instance
> on the server before shutting down the operating system?
If the db is closed manually (shutdown immediate by svrmgrl), there isn't any crash recovery at following startup; the problem shows only if the db is closed by windows services, even if same services are correctly
set up.

>
> An 'instance crash recovery' is not necessarily a bad thing and doesn't
> mean a corrupted database. It mainly means that the shutdown was abnormal
> and that the transaction cleanup that would happen as a normal shutdown is
> being performed during startup.

If the "unclean" shutdown is performed at the "wrong" time the db will probably be corrupted at following startup

>
> However, I encourage you to look at upgrading to 8i or 10g as soon as
> feasible since 9i isw no longer supported. (And that implies your access
> to metalink for 8i is wasted - as you found out looking for patches.)

I've about 40 DBs to manage (from 7.x to 10g) and not all applications can be easily re-written to work with a 9i or upper; moreover this fact justifies metalink access.

Thanks,

Alessandro

-- 
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG