Veritas-bu

[Veritas-bu] Expert advice

2005-10-20 14:25:09
Subject: [Veritas-bu] Expert advice
From: Bill_Jorgensen AT csgsystems DOT com (Jorgensen, Bill)
Date: Thu, 20 Oct 2005 12:25:09 -0600
This is a multi-part message in MIME format.

------_=_NextPart_001_01C5D5A3.9AB09E7B
Content-Type: text/plain;
        charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Mark:
=20
I really appreciate the feedback. Thanks for taking the time to help.
=20
Bill
=20
----------------------------------
   Bill Jorgensen
   CSG Systems, Inc.
   (w) 303.200.3282
   (p) 303.947.9733
----------------------------------
=20

________________________________

From: Mark.Donaldson AT cexp DOT com [mailto:Mark.Donaldson AT cexp DOT com]=20
Sent: Thursday, October 20, 2005 12:10 PM
To: Jorgensen, Bill; veritas-bu AT mailman.eng.auburn DOT edu
Subject: RE: [Veritas-bu] Expert advice


To optimize for restorals, I'd:
=20
1. Turn off multiplexing - mpx doesn't hurt you for single-file restore
but it can cost big on whole server/filesystem restores.
=20
2. Use fragments on your storage units.  I use 4096M per fragment for my
tapes.  This allows faster search-to-file ability at the cost of a
larger catalog.
=20
3. Use multiple volume pools and group sets of backups (and therefore
policies) to each pool.  This would allow you to restore multiple
systems with reduced risk of having to wait on a tape to become
available.  Max value would be as many pools as you have BCP-needed
servers.  Sane level would be number of pools that you have tape drives.
Penalty would be in the sanity cost of managing more policies, pools,
etc. and in capacity issues with so many tapes with free space on the
ends.
=20
That's what pops to mind at first, anyway.
=20
-M

        -----Original Message-----
        From: veritas-bu-admin AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu]On Behalf Of Jorgensen,
Bill
        Sent: Thursday, October 20, 2005 11:11 AM
        To: veritas-bu AT mailman.eng.auburn DOT edu
        Subject: [Veritas-bu] Expert advice
=09
=09
        Guys:
        =20
        Since I am a system administrator that doubles (when time
permits) to do NBU I would like to ask for your advice. Sorry for the
novel, but I would greatly appreciate your input. We are presently
running 5.1MP1on Solaris 9. I have been asked to try and optimize our
configuration for restorals (BCP/DR-related) not backups.
        =20
        The problem right now is that with our current config we can
have as many as 20-30 clients backed up to a single tape. This occurs
with our diffs mostly. Although I have seen multiple clients associated
with a single tape with fulls. When we have multiple clients backed up
on to a single tape it proves to be a significant bottleneck with
restorals as they apply to our BCP/DR plan.
        =20
        We have 49 clients that we have slated for recovery within our
plan. All of these clients are part of one backup policy. We do
multiplex, but we have the setting at 2 for right now. Some of these
servers have either Oracle or DB2. The DB backups are handled with a
separate policy. However, we use one volume pool for media for all
backups regardless of their affiliation or nonaffiliation with BCP/DR.
        =20
        The thought is to create a separate pool for each of the
policies. Do you think that could help us? I would love to hear how you
think we could minimize the number of clients associated with one tape
to make the restoral process less of a bottleneck. As a matter of fact,
I would love to hear anything you have that would help optimize a config
for restorals.
        =20
        Thanks in advance,
        =20
        Bill Jorgensen


------_=_NextPart_001_01C5D5A3.9AB09E7B
Content-Type: text/html;
        charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<META content=3D"MSHTML 6.00.2900.2722" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D245342418-20102005><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Mark:</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D245342418-20102005><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D245342418-20102005><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>I really appreciate the feedback. Thanks for =
taking the=20
time to help.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D245342418-20102005><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D245342418-20102005><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Bill</FONT></SPAN></DIV>
<DIV>&nbsp;</DIV>
<DIV align=3Dleft><FONT face=3DArial=20
size=3D2>----------------------------------</FONT></DIV>
<DIV align=3Dleft><FONT face=3DArial size=3D2>&nbsp;&nbsp; Bill =
Jorgensen</FONT></DIV>
<DIV align=3Dleft><FONT face=3DArial size=3D2>&nbsp;&nbsp; CSG Systems,=20
Inc.</FONT></DIV>
<DIV align=3Dleft><FONT face=3DArial size=3D2>&nbsp;&nbsp; (w)=20
303.200.3282</FONT></DIV>
<DIV align=3Dleft><FONT face=3DArial size=3D2>&nbsp;&nbsp; (p)=20
303.947.9733</FONT></DIV>
<DIV align=3Dleft><FONT face=3DArial=20
size=3D2>----------------------------------</FONT></DIV>
<DIV>&nbsp;</DIV><BR>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
<HR tabIndex=3D-1>
<FONT face=3DTahoma size=3D2><B>From:</B> Mark.Donaldson AT cexp DOT com=20
[mailto:Mark.Donaldson AT cexp DOT com] <BR><B>Sent:</B> Thursday, October 20, =
2005=20
12:10 PM<BR><B>To:</B> Jorgensen, Bill;=20
veritas-bu AT mailman.eng.auburn DOT edu<BR><B>Subject:</B> RE: [Veritas-bu] =
Expert=20
advice<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =
size=3D2>To=20
optimize for restorals, I'd:</FONT></SPAN></DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =
size=3D2>1.=20
Turn off multiplexing - mpx doesn't hurt you for single-file restore but =
it can=20
cost big on whole server/filesystem restores.</FONT></SPAN></DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =
size=3D2>2. Use=20
fragments on your storage units.&nbsp; I use 4096M&nbsp;per fragment for =
my=20
tapes.&nbsp; This allows faster search-to-file ability at the cost of a =
larger=20
catalog.</FONT></SPAN></DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =
size=3D2>3. Use=20
multiple volume pools and group sets of backups (and therefore policies) =
to each=20
pool.&nbsp; This would allow you to restore multiple systems with =
reduced risk=20
of having to wait on a tape to become available.&nbsp; Max value would =
be as=20
many pools as you have BCP-needed servers.&nbsp; Sane level would be =
number of=20
pools that you have tape drives.&nbsp; Penalty would be in the sanity =
cost of=20
managing more policies, pools, etc. and in capacity issues with so many =
tapes=20
with free space on the ends.</FONT></SPAN></DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =
size=3D2>That's=20
what pops to mind at first, anyway.</FONT></SPAN></DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D367565917-20102005><FONT face=3DArial color=3D#0000ff =

size=3D2>-M</FONT></SPAN></DIV>
<BLOCKQUOTE dir=3Dltr style=3D"MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader dir=3Dltr align=3Dleft><FONT =
face=3DTahoma=20
  size=3D2>-----Original Message-----<BR><B>From:</B>=20
  veritas-bu-admin AT mailman.eng.auburn DOT edu=20
  [mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu]<B>On Behalf Of =
</B>Jorgensen,=20
  Bill<BR><B>Sent:</B> Thursday, October 20, 2005 11:11 AM<BR><B>To:</B> =

  veritas-bu AT mailman.eng.auburn DOT edu<BR><B>Subject:</B> [Veritas-bu] =
Expert=20
  advice<BR><BR></FONT></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2>Guys:</FONT></SPAN></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial =
size=3D2>Since I am a=20
  system administrator that doubles (when time permits) to do NBU I =
would like=20
  to ask for your advice. Sorry for the novel, but I would greatly =
appreciate=20
  your input. </FONT></SPAN><SPAN class=3D346391916-20102005><FONT =
face=3DArial=20
  size=3D2>We are presently running 5.1MP1on Solaris 9. I have been =
asked to try=20
  and optimize our configuration for restorals (BCP/DR-related) not=20
  backups.</FONT></SPAN></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial size=3D2>The =
problem right=20
  now is that with our current config we can have as many as 20-30 =
clients=20
  backed up to a single tape. This occurs with our diffs mostly. =
Although I have=20
  seen multiple clients associated with a single tape with fulls. When =
we have=20
  multiple clients backed up on to a single tape it proves to be a =
significant=20
  bottleneck with restorals as they apply to our BCP/DR=20
plan.</FONT></SPAN></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial size=3D2>We =
have 49 clients=20
  that we have slated for recovery within our plan. All of these clients =
are=20
  part of one backup policy. We do multiplex, but we have the setting at =
2 for=20
  right now. Some of these servers have either Oracle or DB2. The DB =
backups are=20
  handled with a separate policy. However, we use one volume pool for =
media for=20
  all backups regardless of their affiliation or nonaffiliation with=20
  BCP/DR.</FONT></SPAN></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial size=3D2>The =
thought is to=20
  create a separate pool for each of the policies. Do you think that =
could help=20
  us?&nbsp;I would love to hear how you think we could minimize the =
number of=20
  clients associated with one tape to make the restoral process less of =
a=20
  bottleneck. As a matter of fact, I would love to hear anything you =
have that=20
  would help optimize a config for restorals.</FONT></SPAN></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial =
size=3D2>Thanks in=20
  advance,</FONT></SPAN></DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial=20
  size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV><SPAN class=3D346391916-20102005><FONT face=3DArial size=3D2>Bill =

  Jorgensen</FONT></SPAN></DIV></BLOCKQUOTE></BODY></HTML>

------_=_NextPart_001_01C5D5A3.9AB09E7B--

<Prev in Thread] Current Thread [Next in Thread>