[Veritas-bu] question about software compression
2003-03-27 13:39:16
Subject: |
[Veritas-bu] question about software compression |
From: |
Fabbro.Andrew AT cnf DOT com (Fabbro, Andrew P) |
Date: |
Thu, 27 Mar 2003 10:39:16 -0800 |
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_01C2F490.2B9CEE7D
Content-Type: text/plain
In my experiments, client compression or tape-drive compression is never as
good as bzip2/gzip/compress. That make sense since the client/tape-drive
compression is stream compression - it's only looking at 4K or 64K or
whatever at a time, versus bzip2/gzip/compress which can chew away on the
whole file.
-----Original Message-----
From: Donaldson, Mark [mailto:Mark.Donaldson AT experianems DOT com]
Sent: Thursday, March 27, 2003 10:26 AM
To: 'Marvin Blackburn'; Veritas--ListServer (E-mail)
Subject: RE: [Veritas-bu] question about software compression
What kind of SCSI bus? Differential? Two drives per bus should be OK with
diff. I wonder if HP is giving you excuses for poor firmware? Be sure to
check those drive firmware versions.
Software compression is SSSSlllllooooowwwww and puts a measurable effect on
a client if you usually use multiple streams. In my testing, it seems the
compression algorithm isn't that tight, a compressed image on disk will
recompress with "compress" or "gzip" very nicely which shows the algorithm
is pretty loose. I don't know how well it compares to tape compression
rates, though.
Avoid software compression if you can, IMO.
-M
-----Original Message-----
From: Marvin Blackburn [mailto:mblackburn AT glenraven DOT com
<mailto:mblackburn AT glenraven DOT com> ]
Sent: Thursday, March 27, 2003 11:08 AM
To: Veritas--ListServer (E-mail)
Subject: [Veritas-bu] question about software compression
We may have to turn off hardware compression on our DLT8000 drives and move
the compression to NBU.
We currently run 3.4.3 against a P3000 ATL Library. We have two drives on
the same scsi bus and
HP is telling us that this may saturate the bus and cause hardware errors.
What kind of performance/time/media consequences should I expect.
------------------
Marvin Blackburn
Systems Administrator
Glen Raven
"He's no failure. He's not dead yet" --William Lloyd George
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
<http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu>
------_=_NextPart_001_01C2F490.2B9CEE7D
Content-Type: text/html
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii">
<TITLE>Message</TITLE>
<META content="MSHTML 6.00.2800.1126" name=GENERATOR></HEAD>
<BODY>
<DIV><SPAN class=614593718-27032003><FONT face=Arial color=#0000ff size=2>In my
experiments, client compression or tape-drive compression is never as good as
bzip2/gzip/compress. That make sense since the client/tape-drive
compression is stream compression - it's only looking at 4K or 64K or whatever
at a time, versus bzip2/gzip/compress which can chew away on the whole
file.</FONT></SPAN></DIV>
<DIV><SPAN class=614593718-27032003></SPAN><SPAN class=614593718-27032003><FONT
face=Arial color=#0000ff size=2></FONT></SPAN> </DIV>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
<DIV></DIV>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><FONT
face=Tahoma size=2>-----Original Message-----<BR><B>From:</B> Donaldson, Mark
[mailto:Mark.Donaldson AT experianems DOT com] <BR><B>Sent:</B> Thursday,
March 27,
2003 10:26 AM<BR><B>To:</B> 'Marvin Blackburn'; Veritas--ListServer
(E-mail)<BR><B>Subject:</B> RE: [Veritas-bu] question about software
compression<BR><BR></FONT></DIV>
<P><FONT size=2>What kind of SCSI bus? Differential? Two drives
per bus should be OK with diff. I wonder if HP is giving you excuses
for
poor firmware? Be sure to check those drive firmware
versions.</FONT></P>
<P><FONT size=2>Software compression is SSSSlllllooooowwwww and puts a
measurable effect on a client if you usually use multiple streams. In
my
testing, it seems the compression algorithm isn't that tight, a compressed
image on disk will recompress with "compress" or "gzip" very nicely which
shows the algorithm is pretty loose. I don't know how well it compares
to tape compression rates, though.</FONT></P>
<P><FONT size=2>Avoid software compression if you can, IMO.</FONT> <BR><FONT
size=2>-M</FONT> </P>
<P><FONT size=2>-----Original Message-----</FONT> <BR><FONT size=2>From:
Marvin Blackburn [<A
href="mailto:mblackburn AT glenraven DOT com">mailto:mblackburn AT glenraven
DOT com</A>]</FONT>
<BR><FONT size=2>Sent: Thursday, March 27, 2003 11:08 AM</FONT> <BR><FONT
size=2>To: Veritas--ListServer (E-mail)</FONT> <BR><FONT size=2>Subject:
[Veritas-bu] question about software compression</FONT> </P><BR>
<P><FONT size=2>We may have to turn off hardware compression on our DLT8000
drives and move</FONT> <BR><FONT size=2>the compression to NBU.</FONT>
<BR><FONT size=2>We currently run 3.4.3 against a P3000 ATL Library. We
have two drives on</FONT> <BR><FONT size=2>the same scsi bus and</FONT>
<BR><FONT size=2>HP is telling us that this may saturate the bus and cause
hardware errors.</FONT> <BR><FONT size=2>What kind of performance/time/media
consequences should I expect.</FONT> </P><BR><BR>
<P><FONT size=2>------------------</FONT> <BR><FONT size=2>Marvin
Blackburn</FONT> <BR><FONT size=2>Systems Administrator</FONT> <BR><FONT
size=2>Glen Raven</FONT> <BR><FONT size=2>"He's no failure. He's not
dead yet" --William Lloyd George</FONT> </P>
<P><FONT size=2>_______________________________________________</FONT>
<BR><FONT size=2>Veritas-bu maillist -
Veritas-bu AT mailman.eng.auburn DOT edu</FONT> <BR><FONT size=2><A
href="http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu"
target=_blank>http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu</A></FONT>
</P></BLOCKQUOTE></BODY></HTML>
------_=_NextPart_001_01C2F490.2B9CEE7D--
|
|
|