Re: LOG_ARCHIVE_DEST_n

The usual rule I have for this is that 'alter session' applies to one session, 'alter system' affects everybody. So yes, it does seem inconsistent.

Leaping to the case for the defence, however: I believe that a query of v$session would indicate a whole batch of sessions at the top of the list (serial number is usually 1) which belong to the background processes. And if ARCH is a background process (and I think we can agree that it is!), then it too must have a session listed there. And since you are seeking to control how ARCH deals with archiving destinations, you really are only affecting that one session... sort of.

Yeah, I know... it's lame, but it's the best I can come up with.

Regards
HJR "Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote in message news:982927632.10850.2.nnrp-10.9e984b29_at_news.demon.co.uk...
>
> I understand why Buck should be puzzled,
> it looks odd to me as well.
>
> If I issue:
>
> alter session set sort_area_size = 10485760;
>
> it doesn't affect any other session, so why should:
>
> alter session set log_archive_dest_3 = 'something'
>
> make the archive process decide that it should
> be affected by something set in my session.
>
>
>
> I don't think either of us would disagree with your
> argument for being able to issue
>
> alter SYSTEM set log_archive_dest_3 = 'yyyy'
>
>
>
> --
> Jonathan Lewis
> Yet another Oracle-related web site: http://www.jlcomp.demon.co.uk
>
> Practical Oracle 8i: Building Efficient Databases
> Publishers: Addison-Wesley
>
> Reviews at: http://www.jlcomp.demon.co.uk/book_rev.html
>
>
>
> Howard J. Rogers wrote in message <3a90d8b7$1_at_news.iprimus.com.au>...
> >You can defer or enable archiving to a destination at the session
 level -and
> >it makes sense to do so. For example, log_archive_dest_3 is mandatory,
 and
> >becomes full.... that means the entire system will hang when ARCH can't
> >successfully archive there, and the online logs loop round.
> >
> >Hence, the DBA fires up a session, defers archiving to that destination,
 and
> >the database springs back into life. DBA then fixes up the underlying
> >problem, and re-enables archiving there.
> >
> >Regards
> >HJR
> >
> >
> >"Buck Turgidson" <remove-me_jc_va_at_hotmail.com> wrote in message
> >news:bkWj6.4171$TD1.318685_at_bgtnsc05-news.ops.worldnet.att.net...
> >> I am reading up on 8i recovery. It says that the aforementioned
 parameter
 is
> >> session-modifiable. A query against V$PARAMETER confirms this. Maybe
 I
 don't have
> >> the whole context, but this doesn't make sense to me. I could
 understand
 being
> >> system-modifiable, but in what context is it session-modifiable?
> >>
> >> Thanks to anyone who could clear up my confusion.
> >>
> >>
> >
> >
>
>
Received on Fri Feb 23 2001 - 17:24:03 CST