|
[Veritas-bu] Expert advice
2005-10-20 14:09:35
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_01C5D5A1.6CDAD9A0
Content-Type: text/plain;
charset="iso-8859-1"
To optimize for restorals, I'd:
1. Turn off multiplexing - mpx doesn't hurt you for single-file restore but
it can cost big on whole server/filesystem restores.
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.
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.
That's what pops to mind at first, anyway.
-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
Guys:
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.
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.
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.
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.
Thanks in advance,
Bill Jorgensen
------_=_NextPart_001_01C5D5A1.6CDAD9A0
Content-Type: text/html;
charset="iso-8859-1"
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2800.1515" name=GENERATOR></HEAD>
<BODY>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff size=2>To
optimize for restorals, I'd:</FONT></SPAN></DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff size=2>1.
Turn off multiplexing - mpx doesn't hurt you for single-file restore but it can
cost big on whole server/filesystem restores.</FONT></SPAN></DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff size=2>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.</FONT></SPAN></DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff size=2>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.</FONT></SPAN></DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2>That's
what pops to mind at first, anyway.</FONT></SPAN></DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=367565917-20102005><FONT face=Arial color=#0000ff
size=2>-M</FONT></SPAN></DIV>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
<DIV class=OutlookMessageHeader dir=ltr align=left><FONT face=Tahoma
size=2>-----Original Message-----<BR><B>From:</B>
veritas-bu-admin AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu]<B>On Behalf Of
</B>Jorgensen,
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
advice<BR><BR></FONT></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2>Guys:</FONT></SPAN></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial size=2>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. </FONT></SPAN><SPAN class=346391916-20102005><FONT face=Arial
size=2>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.</FONT></SPAN></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial size=2>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.</FONT></SPAN></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial size=2>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.</FONT></SPAN></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial size=2>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.</FONT></SPAN></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial size=2>Thanks in
advance,</FONT></SPAN></DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=346391916-20102005><FONT face=Arial size=2>Bill
Jorgensen</FONT></SPAN></DIV></BLOCKQUOTE></BODY></HTML>
------_=_NextPart_001_01C5D5A1.6CDAD9A0--
|
|
|
