Networker

Re: [Networker] advance file type question

2005-09-30 02:04:37
Subject: Re: [Networker] advance file type question
From: Peter Viertel <Peter.Viertel AT MACQUARIE DOT COM>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Fri, 30 Sep 2005 16:03:25 +1000
I just went though this exercise recently, although I left sol10 /zfs
aside for now.

Tip #1: use solaris 9 rather that 8.

VxFS on vxvm is probably the best solution, but if you don't want to pay
for vxfs, and you are using solaris9 then the constraints are as
follows:


Vxvm can create volumes > 1TB, but you can only put ufs on a vxvm volume
up to 2TB.

If you use SVM then you can create ufs filesystems of many terabytes - I
tested an 8TB one with no problems, although I currently use 2.5TB
filesystems.

In solaris 9, ufs logging is on by default, and I have not yet had to do
a full fsck.

Another thing to watch is LUN size. If the LUNs are over 1TB then the
classic solaris vtoc labels cant be used and you will get new style EFI
labels which are great for SVM, but can't be initialised into vxvm at
least as far as version 4.0 is concerned.

Sol9 SVM can handle redundant paths and relocation reasonably well.
Though its not as convenient as vxvm disk groups yet.

-----Original Message-----
From: Legato NetWorker discussion [mailto:NETWORKER AT listserv.temple DOT edu]
On Behalf Of T. S. Kimball
Sent: Friday, 30 September 2005 3:20 PM
To: NETWORKER AT listserv.temple DOT edu
Subject: Re: [Networker] advance file type question

I've actually opened a Sun case regarding this last year (wanted info so
we could configure our BladeStore correctly).  Quick answers:

UFS on Solaris limits a volume (LUN) limit of 1 TB.  fsck takes
*forever* on these, at least on Solaris 8.

Larger volumes are available if you want to goto either ZFS (their new
disk format, Sol 10 or Sol 9 /w backported pkgs) or use Veritas
VxVM/VxFS to build larger volumes once the 1 TB LUNs are presented to
Solaris.

We chose Veritas running on Solaris 8, in part due to other things such
as Networker Server support in 9/10, AlphaStor support in 9/10, and we
felt that Solaris 10 and ZFS is too new for us to risk the backup
environment on.

BladeStore is a bunch of 3+1 RAID 5's, with three 1 TB LUNS and one 0.5
TB LUN in each RAID group.

For Veritas, we concat the LUNs and create Adv_file devices of 12-16 TB
each  (DBO Tier 5 license).  Need to newfs the vxfs filesystem with
largefiles support but otherwise works well.

This also gives the added advantage that I can Deport a set of adv_file
LUNs and re-present/Import them to a Storage Node when I need to
re-balance devices.  [Note, I've not tested yet, but should work.]


--TSK


-----=====-----
Timothy Kimball          tkimball AT brass DOT com
-----=====-----

To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems wit this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or via RSS at
http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER



NOTICE
This e-mail and any attachments are confidential and may contain copyright 
material of Macquarie Bank or third parties. If you are not the intended 
recipient of this email you should not read, print, re-transmit, store or act 
in reliance on this e-mail or any attachments, and should destroy all copies of 
them. Macquarie Bank does not guarantee the integrity of any emails or any 
attached files. The views or opinions expressed are the author's own and may 
not reflect the views or opinions of Macquarie Bank.

To sign off this list, send email to listserv AT listserv.temple DOT edu and 
type "signoff networker" in the
body of the email. Please write to networker-request AT listserv.temple DOT edu 
if you have any problems
wit this list. You can access the archives at 
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER