Well with a large datazone and media db with many clients the best would be to
set all your backup groups to no index save. This will prevent each client in
the group from doing the index right after that client completes its backup.
Put the backup server in a group schedule standalone and scheduled through
networker. What this will do is prevent a bootstrap backup happening after
every group finishes. I even have a schedule assigned to that group so the will
skip every day but Sunday. Put all your clients in a index group. Create the
pool and assign the group to that pool.
Run savegrp -O -l full groupname in the AM after your backups complete or close
to it. This will backup all indexes full and the server bootstrap. You can even
limit everything to one tape by using the -N flag with say 10. Have your drive
parallelism set to 10. All indexes and bootstrap on one tape most days. You do
not have to do the full on the indexes everyday if you so choose.
EDWARD COTY
LEAD STORAGE ENGINEER, LCNA, NACP
WORK - 973-533-2098
CELL - 973-296-0918
EDWARD.COTY AT AIG DOT COM
-----Original Message-----
From: EMC NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU] On
Behalf Of Bruce Breidall
Sent: Monday, July 14, 2008 4:06 PM
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: [Networker] Bootstrap and Index Question
I was wondering what the best practice might be for storing index entries.
I think the media pool store index entries dictates if indexes are saved.
In our case, the pools have that parameter set, and in the logs I see that the
indexes are saved to the tape after the backups for the that group associated
to that pool. Finally, I see that the bootstrap is written, which also includes
the media DB.
It looks like the media DB is frozen until that job is done. If there are many
groups, and many pools, doesn't this cause a lot of delay and overhead in the
overall scheme of things?
Is it better to remove the store index entries on the media pool, and manually
request a special savegrp -O -G (SGO) once or several times a day? If you are
performing DR cloning, then the last thing you would want to do is get the SGO
backup, and eject that tape, but that seems like a catch 22 if you are
depending on that tape for onsite restores and browsing capability via the
indexes. What are the tradeoffs here?
Alternatively, if you do the SGO backup, and take a clone of that, then you are
automatically not current again on the clone. Another catch 22.
Is there a best practice with this? I don't see any problems with the way I am
doing this, other than the media DB shutting down for the duration of the
bootstrap save, which can't be helping other concurrent activity.
Is there any official documents about this? The DR manual is just about
worthless in discussing and explaining these finer points. I have been
searching the archives, but there is so much there to look at it and try to
narrow down.
Thanks.
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems with this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or via RSS at
http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems with this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
|
