This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_001_01C2341E.78ACF7E0
Content-Type: text/plain
Try setting your SIZE_DATA_BUFFERS equal to NET_BUFFER_SZ, restart NB and
see how it goes. There's notes somewhere in those docs that if these don't
match then performance will suck (I think it matches input sizes to output
sizes).
Also, is you DLT7K writing in 7000 mode or is it dumbing down to an earlier
density setting?
-M
-----Original Message-----
From: Jeff Kennedy [mailto:jlkennedy AT amcc DOT com]
Sent: Thursday, July 25, 2002 2:57 PM
To: Donaldson Mark
Cc: NBU List
Subject: Re: [Veritas-bu] duplication
Here's what I've done:
NET_BUFFER_SZ = 262144
NUMBER_DATA_BUFFERS = 16
SIZE_DATA_BUFFERS = 65536
All entries in /etc/system are verified as correct and all of them are
present according to the tech doc. Calculated shmem is based on new
values for SIZE_DATA_BUFFERS and NUMBER_DATA_BUFFERS, 4 drives, and a 4
mpx per drive.
I rebooted the server and I still get <1mb/sec on a duplication job. A
single unix client network backup runs at >8mb/sec so it's not tcp stack
or nic problems. And again, there is no mpx issue since this data is
ndmp.
Server is:
Sun E250, 2 X 400MHz, 2gb ram, /usr/openv on A51000 striped/mirrored
aray, tape drives are HVD scsi direct attached.
Anyone have any ideas? Case open with Veritas but no answer yet.
Thanks.
~JK
> "Donaldson, Mark" wrote:
>
> Check your buffer settings. This, and a demux source, made a
> difference for me.
>
> See: http://seer.support.veritas.com/docs/183702.htm
> and: http://seer.support.veritas.com/docs/243197.htm
>
> FWIW: I get much better performance on my DLT7k (ultra-diff) than your
> 1mb/sec rates.
>
> -M
>
> -----Original Message-----
> From: Jeff Kennedy [mailto:jlkennedy AT amcc DOT com]
> Sent: Wednesday, July 24, 2002 11:56 AM
> To: NBU List
> Subject: [Veritas-bu] duplication
>
> I don't duplicate much but does it always run slow? I am seeing an
> average read/write of 1mb/sec for a duplication on dlt7k drives
> attached
> to the master server. The primary copy is an ndmp backup so there's
> no
> multiplexing involved.
> --
> =====================
> Jeff Kennedy
> Unix Administrator
> AMCC
> jlkennedy AT amcc DOT com
> _______________________________________________
> Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
--
=====================
Jeff Kennedy
Unix Administrator
AMCC
jlkennedy AT amcc DOT com
------_=_NextPart_001_01C2341E.78ACF7E0
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3DUS-ASCII">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2653.12">
<TITLE>RE: [Veritas-bu] duplication</TITLE>
</HEAD>
<BODY>
<P><FONT SIZE=3D2>Try setting your SIZE_DATA_BUFFERS equal to =
NET_BUFFER_SZ, restart NB and see how it goes. There's notes =
somewhere in those docs that if these don't match then performance will =
suck (I think it matches input sizes to output sizes).</FONT></P>
<P><FONT SIZE=3D2>Also, is you DLT7K writing in 7000 mode or is it =
dumbing down to an earlier density setting?</FONT>
</P>
<P><FONT SIZE=3D2>-M</FONT>
</P>
<P><FONT SIZE=3D2>-----Original Message-----</FONT>
<BR><FONT SIZE=3D2>From: Jeff Kennedy [<A =
HREF=3D"mailto:jlkennedy AT amcc DOT com">mailto:jlkennedy AT amcc DOT
com</A>]</FONT>=
<BR><FONT SIZE=3D2>Sent: Thursday, July 25, 2002 2:57 PM</FONT>
<BR><FONT SIZE=3D2>To: Donaldson Mark</FONT>
<BR><FONT SIZE=3D2>Cc: NBU List</FONT>
<BR><FONT SIZE=3D2>Subject: Re: [Veritas-bu] duplication</FONT>
</P>
<BR>
<P><FONT SIZE=3D2>Here's what I've done:</FONT>
</P>
<P><FONT SIZE=3D2>NET_BUFFER_SZ =3D 262144</FONT>
<BR><FONT SIZE=3D2>NUMBER_DATA_BUFFERS =3D 16</FONT>
<BR><FONT SIZE=3D2>SIZE_DATA_BUFFERS =3D 65536</FONT>
</P>
<P><FONT SIZE=3D2>All entries in /etc/system are verified as correct =
and all of them are</FONT>
<BR><FONT SIZE=3D2>present according to the tech doc. Calculated =
shmem is based on new</FONT>
<BR><FONT SIZE=3D2>values for SIZE_DATA_BUFFERS and =
NUMBER_DATA_BUFFERS, 4 drives, and a 4</FONT>
<BR><FONT SIZE=3D2>mpx per drive.</FONT>
</P>
<P><FONT SIZE=3D2>I rebooted the server and I still get <1mb/sec on =
a duplication job. A</FONT>
<BR><FONT SIZE=3D2>single unix client network backup runs at =
>8mb/sec so it's not tcp stack</FONT>
<BR><FONT SIZE=3D2>or nic problems. And again, there is no mpx =
issue since this data is</FONT>
<BR><FONT SIZE=3D2>ndmp.</FONT>
</P>
<P><FONT SIZE=3D2>Server is:</FONT>
</P>
<P><FONT SIZE=3D2>Sun E250, 2 X 400MHz, 2gb ram, /usr/openv on A51000 =
striped/mirrored</FONT>
<BR><FONT SIZE=3D2>aray, tape drives are HVD scsi direct =
attached.</FONT>
</P>
<P><FONT SIZE=3D2>Anyone have any ideas? Case open with Veritas =
but no answer yet. </FONT>
<BR><FONT SIZE=3D2>Thanks.</FONT>
</P>
<P><FONT SIZE=3D2>~JK</FONT>
</P>
<P><FONT SIZE=3D2>> "Donaldson, Mark" wrote:</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> Check your buffer settings. This, and a =
demux source, made a</FONT>
<BR><FONT SIZE=3D2>> difference for me.</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> See: <A =
HREF=3D"http://seer.support.veritas.com/docs/183702.htm"; =
TARGET=3D"_blank">http://seer.support.veritas.com/docs/183702.htm</A></F=
ONT>
<BR><FONT SIZE=3D2>> and: <A =
HREF=3D"http://seer.support.veritas.com/docs/243197.htm"; =
TARGET=3D"_blank">http://seer.support.veritas.com/docs/243197.htm</A></F=
ONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> FWIW: I get much better performance on my DLT7k =
(ultra-diff) than your</FONT>
<BR><FONT SIZE=3D2>> 1mb/sec rates.</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> -M</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> -----Original Message-----</FONT>
<BR><FONT SIZE=3D2>> From: Jeff Kennedy [<A =
HREF=3D"mailto:jlkennedy AT amcc DOT com">mailto:jlkennedy AT amcc DOT
com</A>]</FONT>=
<BR><FONT SIZE=3D2>> Sent: Wednesday, July 24, 2002 11:56 AM</FONT>
<BR><FONT SIZE=3D2>> To: NBU List</FONT>
<BR><FONT SIZE=3D2>> Subject: [Veritas-bu] duplication</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> I don't duplicate much but does it always run =
slow? I am seeing an</FONT>
<BR><FONT SIZE=3D2>> average read/write of 1mb/sec for a duplication =
on dlt7k drives</FONT>
<BR><FONT SIZE=3D2>> attached</FONT>
<BR><FONT SIZE=3D2>> to the master server. The primary copy is =
an ndmp backup so there's</FONT>
<BR><FONT SIZE=3D2>> no</FONT>
<BR><FONT SIZE=3D2>> multiplexing involved.</FONT>
<BR><FONT SIZE=3D2>> --</FONT>
<BR><FONT SIZE=3D2>> =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D</FONT>
<BR><FONT SIZE=3D2>> Jeff Kennedy</FONT>
<BR><FONT SIZE=3D2>> Unix Administrator</FONT>
<BR><FONT SIZE=3D2>> AMCC</FONT>
<BR><FONT SIZE=3D2>> jlkennedy AT amcc DOT com</FONT>
<BR><FONT SIZE=3D2>> =
_______________________________________________</FONT>
<BR><FONT SIZE=3D2>> Veritas-bu maillist - =
Veritas-bu AT mailman.eng.auburn DOT edu</FONT>
<BR><FONT SIZE=3D2>> <A =
HREF=3D"http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu"; =
TARGET=3D"_blank">http://mailman.eng.auburn.edu/mailman/listinfo/veritas=
-bu</A></FONT>
</P>
<P><FONT SIZE=3D2>-- </FONT>
<BR><FONT =
SIZE=3D2>=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
</FONT>
<BR><FONT SIZE=3D2>Jeff Kennedy</FONT>
<BR><FONT SIZE=3D2>Unix Administrator</FONT>
<BR><FONT SIZE=3D2>AMCC</FONT>
<BR><FONT SIZE=3D2>jlkennedy AT amcc DOT com</FONT>
</P>
</BODY>
</HTML>
------_=_NextPart_001_01C2341E.78ACF7E0--
|
