Re: Is there a limit of rows in V$ARCHIVED_LOG

You know, just when I'm tempted to sigh at yet another question about case insensitivity, I realise why I like this group so much: anyone and his dog can *learn* so much from here. OK, I'll surrender on this one... provided you do what Jonathan asked, and change keep_time to be 21 days or there abouts, and then do your tests!

See you in February!!

Regards
HJR "Brian Peasland" <peasland_at_edcmail.cr.usgs.gov> wrote in message news:3A6EF404.6D5991CB_at_edcmail.cr.usgs.gov...
> Howard,
>
> I did some more checking on this...
>
> If I look at the entries in V$ARCHIVED_LOG, I can see that the entries
> go back exactly 7 days!! This matches my CONTROL_FILE_RECORD_KEEP_TIME
> parameter. If I examine my control file, I can see that I have
> MAXLOGHISTORY set to 3630. But my V$ARCHIVED_LOG view only contains 1524
> entries for those 7 days. So it seems to me that the
> CONTROL_FILE_RECORD_KEEP_TIME parameter is the factor that determines
> how long entries are kept in V$ARCHIVED_LOG.
>
> More food for thought...
> Brian
>
>
>
> "Howard J. Rogers" wrote:
> >
> > "Brian Peasland" <peasland_at_edcmail.cr.usgs.gov> wrote in message
> > news:3A6DA7C9.CABE38BF_at_edcmail.cr.usgs.gov...
> > > > I may be hopelessly wrong about this, but I could almost swear that
> > > > control_file_record_keep_time only becomes of relevance when using
 RMAN
> > > > without a catalogue, and that the real problem is maxloghistory when
> > > > creating the database.
> > >
> > > Actually, I thought that it was somewhere in the middle....
> > >
> > > I believe that the MAXLOGHISTORY is as you say it to be. But the
> > > CONTROL_FILE_RECORD_KEEP_TIME does *not only* apply to RMAN. I've
 never
> > > used RMAN on any of my databases. But when I migrate a database, I
 will
> > > see my control file expand for the number of days specified in the
 above
> > > parameter. This is due to the control file having to capture more
> > > information than the previous db version. So the control file expands
 in
> > > size. Once the CONTROL_FILE_RECORD_KEEP_TIME number of days has
> > > passed,the control file doesn't have to expand since it can write over
> > > previous information. If one uses RMAN, then RMAN gets information
 from
> > > the control file and it can go back the number of days specified in
 this
> > > parameter. So this parameter must be adjusted if you want RMAN to go
> > > back far enough. IIRC, I believe the default for this parameter is 7
> > > days (Oracle 8i on Solaris).
> > >
> > > Of course, this is just my experiences with a not so obviously
> > > documented parameter and I could be wrong......
> > >
> > > Brian
> > >
> >
> > Hi Brian -well, you're dead right that the default for keep_time is 7
 days.
> > I guess there's likely an interaction between the two things as well.
 Live
> > and Learn, eh!?
> >
> > Regards
> > HJR
> >
> > >
> > > --
> > > ========================================
> > > Brian Peasland
> > > Raytheons Systems at
> > > USGS EROS Data Center
> > > These opinions are my own and do not
> > > necessarily reflect the opinions of my
> > > company!
> > > ========================================
>
> --
> ========================================
> Brian Peasland
> Raytheons Systems at
> USGS EROS Data Center
> These opinions are my own and do not
> necessarily reflect the opinions of my
> company!
> ========================================
Received on Wed Jan 24 2001 - 14:10:18 CST