[Veritas-bu] Fragment Size
2002-05-15 12:50:40
Subject: |
[Veritas-bu] Fragment Size |
From: |
rlh AT lsil DOT com (Richard Hellier) |
Date: |
Wed, 15 May 2002 17:50:40 +0100 |
This is a multi-part message in MIME format.
--------------36657D7FF6B7ED24D8949F07
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
We run with a fragment size of 2G and, AFAIK, backup speed is not
impacted at all (or significantly, at least). The gain comes when
doing
a restore as the process can skip across whole fragments before reaching
the area of the tape containing the material to be restored.
With a zero fragment size (i.e. don't fragment), the tar images
are written as a single monolithic block which is slower to scan
through.
I didn't understand Fabrice's point about end of tape -- I don't
see how running "off the end" is better / worse if using fragments or
not.
Can someone enlighten me on this point?
Cheers,
Richard.
--------------36657D7FF6B7ED24D8949F07
Content-Type: text/x-vcard; charset=us-ascii;
name="rlh.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Richard Hellier
Content-Disposition: attachment;
filename="rlh.vcf"
begin:vcard
n:Hellier;Richard
tel;pager:[+44] 7693-905798
tel;fax:[+44] 1344-413114
tel;work:[+44] 1344-413183
x-mozilla-html:FALSE
org:LSI Logic;GES UNIX Europe
version:2.1
email;internet:rlh AT lsil DOT com
adr;quoted-printable:;;Greenwood House=0D=0ALondon Road;Bracknell;BK;RG12 2UB;GB
x-mozilla-cpt:;0
fn:Richard Hellier
end:vcard
--------------36657D7FF6B7ED24D8949F07--
|
<Prev in Thread] |
Current Thread |
[Next in Thread>
|
- [Veritas-bu] Fragment Size, Carlos Perales
- [Veritas-bu] Fragment Size, Brochart, Fabrice
- [Veritas-bu] Fragment Size, White, Steve
- [Veritas-bu] Fragment Size, Brochart, Fabrice
- [Veritas-bu] Fragment Size, Sparks, Klenton
- [Veritas-bu] Fragment Size, Shahar.Shaynis AT ecitele DOT com
- [Veritas-bu] Fragment Size, Max.Booth AT ubsw DOT com
|
|
|