One interesting method, is to backup to disk, and then use HSM to migrate
the files to tape.
Just a thought.
Bob
-----Original Message-----
From: Joshua Fielden [mailto:jfielden AT excitecorp DOT com]
Sent: Tuesday, November 21, 2000 3:59 PM
To: W. Curtis Preston
Cc: Gustavo Leite de Mendonca Chaves; veritas-bu AT mailman.eng.auburn DOT edu
Subject: Re: [Veritas-bu] How to change the working queue priority on
the fly?
I'm pretty sure the naswer here is 'use vault', but is there a way to use
this setup in a 'holding disk/staging' scenario, ala Networker? (not ala
AMANDA)
JF
On Tue, Nov 21, 2000 at 01:27:33PM -0800, W. Curtis Preston spake unto the
multitudes:
>
> The only real solution, IMHO, is to create a disk-based storage pool that
> can hold 1-3 days worth of logical logs. Backups and recoveries from disk
> are MUCH faster, and it gets around the problem you are describing. Then
> you create a class that backs up that disk-based-storage unit to tape
every
> hour. Then you have a cron job that cleans it out once a day.
>
> This is what we do.
>
> At 05:49 PM 11/20/00 -0200, Gustavo Leite de Mendonca Chaves wrote:
> >Dear admins,
> >
> >I'm facing a schedule problem for which I can't find a good solution
> >so I decided to use your collective wisdom.
> >
> >I'm running NBU 3.2 with patch J0820412 on a Solaris 2.6 master server
> >and I have a Solaris 2.7 as a slave server. Each server is attached
> >to two DLT7000 drives inside a STK 9714 library. I also use the
> >Informix option to backup an Informix database.
> >
> >My problem is that during the peak of the scheduled full/cumulative
> >backups that goes from Friday/21:00:00 until Saturday/12:00:00 more or
> >less all four drives are busy with tapes from the mensal(monthly) and
> >semanal(weekly) pools so that the backups initiated by the Informix
> >database for its logs can't be performed because they want to use
> >tapes from the diario(daily) pool.
> >
> >One solution would be to increase the number of Informix log files so
> >that they couldn't be filled in less than 15 hours but as things are
> >they can be filled in less than 3 hours depending on the amount of
> >work it's doing. I would have to multiply the log space by 10 to be
> >sure it was safe and I can't do it right now.
> >
> >To bypass the problem what I did (and still do) was to define two
> >`Backup Policy' schedules for the Informix-ON-BAR class with
> >complementary windows like so:
> >
> > Schedule Volume Pool Window
> > Default-Policy diario Saturday(12:00) -> Friday(21:00)
> > Default-Weekend mensal Friday(21:00) ->
Saturday(12:00)
> >
> >The intent is to direct the log backups to the diario pool most of the
> >time and to direct them to the mensal pool during the peak scheduled
> >backups when the chance of there being a mensal tape in one of the
> >drives is very high.
> >
> >This kludge has worked more or less. The problem is that when some
> >drives go down or something else happens that cause a delay or a
> >change to the peak period I have to foresee it and manually change the
> >schedules above. This has been very unreliable.
> >
> >What I really wanted was a way to tell the Default Policy schedule to
> >automatically use a tape from another pool if all drives are busy with
> >tapes not from the diario pool. Or else, if I could simply tell NBU
> >that the log backup jobs are so high priority that they should put
> >them on the top of the working queue. The Informix-ON-BAR class is
> >already the most privileged one but this seems to not have the desired
> >effect on the log backups.
> >
> >As a related issue, does anyone know how/if I can list the current
> >working list to see how the queued jobs are sorted and how/if I can
> >change the order myself? I think this would be a very useful feature
> >and would let me solve this particular problem. What do you think?
> >
> >I'd appreciate any help you can give to me.
> >
> >Thanks.
> >
> >Gustavo.
> >_______________________________________________
> >Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
> >http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
>
> ---
> W. Curtis Preston, Principal Consultant at Collective Technologies
> Email: curtis AT colltech DOT com (Best way to contact me)
> Work : 408 452 5555 (Leave a message.)
> Pager: 800 946 4646, pin#1436065 (If urgent.)
>
> Tap into the Collective Intellect (TM): http://www.colltech.com
> Backup & Restore resources: http://www.backupcentral.com
>
> _______________________________________________
> Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
--
"Any man page that includes the words "USE AT YOUR OWN RISK. BEWARE OF DOG.
SLIPPERY WHEN WET" means trouble" - Michael Lucas
Joshua Fielden, Senior Systems Administrator and Backups Team Lead
eXcite@Home, Inc. jfielden AT excitecorp DOT com 650-556-3316
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|
