why did you change fragment size?
It is not necessary and does more harm than good.....
Dean wrote:
> Yep, fragment size is set at 20GB on 3592-E05 drives
>
> On Sat, Oct 18, 2008 at 8:47 AM, oersted
> <netbackup-forum < at > backupcentral.com> wrote:
>
> >
> > do you have fragment size set on your storage units to something other than
> > default?
> >
> >
> > Dean wrote:
> >
> > > I am sitting here at home on a Friday night watching the backups run
> > > (yeah, I have no life), trying to get an idea about how the
> > > performance is going, and trying to help a DBA narrow down a
> > > performance problem. I have 3 backups running at the moment (things
> > > haven't really kicked off yet). Everything was going fine - I was
> > > seeing 60 MB/sec from the main client I am interested in, and about 20
> > > MB/sec from the other two, which is all good.
> > >
> > > Then suddenly everything freezes. It has been about half an hour and
> > > nothing has happened.
> > >
> > > The only error message I see is this one....
> > >
> > > 1224235829 1 2 16 melrhnbu02 0 0 0 *NULL* bptm Error occurred during
> > > initialization. Could not read logging configuration file.
> > >
> > > (that is from a media server which is not involved in the particular
> > > backup I am intersted in, so is probably completely irrelevant)
> > >
> > > I also get lots of this :
> > > bpcd/log.101708:07:43:53.526 [4941] <16> bpcd main: char_count = 7375,
> > > .line_count = 55
> > > bpcd/log.101708:07:43:57.112 [4954] <16> bpcd main: strlen(pProcList) =
> > > 7205
> > >
> > > GRRRR
> > >
> > > It;s extremely frustrating having watched a 100GB filesystem backup
> > > tick up to about 90GB completed at 60 MB/sec, then everything freezes,
> > > when I have a DBA waiting for me to tell him that everything looks OK
> > > from the I/O side of things on his server. But I can't give him a
> > > definitive answer, because EVERYTHING has locked up for me, from all 3
> > > clients that are trying to backup at the moment.
> > >
> > > Oh NetBackup, why must thou smite me?
> > >
> > > Well, I was just about to send this email, and things have suddenly
> > > started moving again. That's life with NBU, I guess. I should be used
> > > to these mysterious lock-ups by now. But it's extremely frustrating,
> > > especially when you're trying to do an I/O performance test.
> > >
> > > So, I guess I'll have to report to the DBA, the first 4 x 20GB
> > > fragments ran at 60 MB/sec, but the last 20GB fragment ran at less
> > > than 1 MB/sec. What is he going to make of that?
> > >
> > > NBU 6.5.2 on RHEL 4. Generally it's pretty stable for me, but random
> > > lock ups like this continue to occur, as they have since I started
> > > with NBU in the 3.4 days.
> > >
> > > Gehhh, just getting out my frustrations. Thanks for listening.
> > > _______________________________________________
> > > Veritas-bu maillist - Veritas-bu < at > mailman.eng.auburn.edu
> > > http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
> > >
> >
> >
> > +----------------------------------------------------------------------
> > |This was sent by solaris < at > cablespeed.com via Backup Central.
> > |Forward SPAM to abuse < at > backupcentral.com.
> > +----------------------------------------------------------------------
> >
> >
> > _______________________________________________
> > Veritas-bu maillist - Veritas-bu < at > mailman.eng.auburn.edu
> > http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
> >
> >
> _______________________________________________
> Veritas-bu maillist - Veritas-bu < at > mailman.eng.auburn.edu
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
+----------------------------------------------------------------------
|This was sent by solaris AT cablespeed DOT com via Backup Central.
|Forward SPAM to abuse AT backupcentral DOT com.
+----------------------------------------------------------------------
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|