Dwayne

Quite possibly right as we are using STK drivers for the drives, was
under the impression it was an OS limitation though.

What I'm not so clear on is that if drives are directly attached to the
filer and in our case zoned out from anything else. The filer actually
does all the writing so how would SIZE_DATA_BUFFERS_NDMP be implemented,
somewhere on the filer itself ?


Dave
-----Original Message-----
From: Dwayne Adams [mailto:dadams AT organic DOT com] 
Sent: 21 August 2007 15:06
To: Clooney, David; Martin, Jonathan; Veritas-bu AT mailman.eng.auburn DOT edu
Subject: RE: [Veritas-bu] NDMP Performance

Hello,

I tested changing the block size using these settings with Onstor filers
and came up empty.  I matched block size changes on the filer with
SIZE_DATA_BUFFERS_NDMP.  Every test of changing the block size resulted
in I/O errors.  In the end NDMP performance seems to be highly dependent
on the performance of the underlying file system you are backing up
along with NDMP throttling to guarantee a certain level of performance
for the users accessing the data during the backup.  I have also read
that you don't have the 64k limit in Windows if you use the VERITAS
drivers.

-Dwayne Adams
-----Original Message-----
From: veritas-bu-bounces AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-bounces AT mailman.eng.auburn DOT edu] On Behalf Of Clooney,
David
Sent: Tuesday, August 21, 2007 6:22 AM
To: Martin, Jonathan; Veritas-bu AT mailman.eng.auburn DOT edu
Subject: Re: [Veritas-bu] NDMP Performance

Good idea,

>>From the unix world myself, however we have overcome the blocksize
limit
on our windows media servers with Microsoft KB907418.

2003 SP1 I presume

Anyhow I think they have now bundled the fix into the next SP release
and is therefore not available for download. I have a copy of the fix if
you need it.

Dave



-----Original Message-----
From: veritas-bu-bounces AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-bounces AT mailman.eng.auburn DOT edu] On Behalf Of Martin,
Jonathan
Sent: 21 August 2007 13:34
To: Veritas-bu AT mailman.eng.auburn DOT edu
Subject: Re: [Veritas-bu] NDMP Performance

I put in a call to Symantec Tech Support on this, given that its not
documented anywhere and the tech indicated this is the write to tape
block size.  So given that your tape drive can handle it, setting this
as high as 256KB can net you performance improvements.  I however am on
Windows, so 64K seems to be it for me.

-Jonathan 

-----Original Message-----
From: Clooney, David [mailto:david.clooney AT bankofamerica DOT com] 
Sent: Tuesday, August 21, 2007 4:31 AM
To: Len Boyle; Martin, Jonathan; Veritas-bu AT mailman.eng.auburn DOT edu
Subject: RE: [Veritas-bu] NDMP Performance

I was under the impression that if the drives are directly attached
there is not a lot that can be done from an NBU point of view.

If you are running a 3 way scenario then the touch file could possibly
make a difference, I seem to remember from the past that if you are in a
3 way setup you can change the way the NDMP mover agent starts with
regard to buffers.

Dave

-----Original Message-----
From: veritas-bu-bounces AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-bounces AT mailman.eng.auburn DOT edu] On Behalf Of Len
Boyle
Sent: 21 August 2007 04:06
To: Martin, Jonathan; Veritas-bu AT mailman.eng.auburn DOT edu
Subject: Re: [Veritas-bu] NDMP Performance

The answer may depend on what device is at the other end of NDMP. 
I was told that with NetApp NDMP that changing the buffer did not change
performance due to the way that the WAFL filesystem handled things. 
It would be interesting to revisit this and see if they had made any
changes to enable larger buffers sizes improve performance. 

Some of the other NDMP devices with different filesystems may show an
improvement.

len

-----Original Message-----
From: veritas-bu-bounces AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-bounces AT mailman.eng.auburn DOT edu] On Behalf Of Martin,
Jonathan
Sent: Monday, August 20, 2007 3:12 PM
To: Veritas-bu AT mailman.eng.auburn DOT edu
Subject: [Veritas-bu] NDMP Performance

Anyone know anything about the SIZE_DATA_BUFFERS_NDMP touch file?  Is
this even still valid?  I can't seem to find anything about it in NBU
documentation, just a reference in the archives of this mailing list.
Is this worth investigating as a performance enhancer for NDMP?

-Jonathan

_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu

_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu



Notice to recipient:
The information in this internet e-mail and any attachments is
confidential and may be privileged. It is intended solely for the
addressee. If you are not the intended addressee please notify the
sender immediately by telephone. If you are not the intended recipient,
any disclosure, copying, distribution or any action taken or omitted to
be taken in reliance on it, is prohibited and may be unlawful.

When addressed to external clients any opinions or advice contained in
this internet e-mail are subject to the terms and conditions expressed
in any applicable governing terms of business or client engagement
letter issued by the pertinent Bank of America group entity.

If this email originates from the U.K. please note that Bank of America,
N.A., London Branch and Banc of America Securities Limited are
authorised and regulated by the Financial Services Authority.

_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu



- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
This email is intended only for the person or entity to which it is
addressed
and may contain information that is privileged, confidential or
otherwise
protected from disclosure. Dissemination, distribution or copying of
this
email or the information herein by anyone other than the intended
recipient,
or an employee or agent responsible for delivering the message to the
intended
recipient, is prohibited. If you have received this email in error,
please
immediately notify us by calling our Help Desk at (415) 581-5552 or
by e-mailing us at helpdesk AT organic DOT com.

_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu