Mark,

I've seen occasions when disks have failed on a TSM server, and a user 
calls to have a single Word document retrieved. You want to be the one 
explaining to the user that the disk on the backupserver has failed, and 
that his/hers document cannot be restored? :=)

Fault tolerant disks are never to save you against a disaster. If so, you 
need to remotely mirror the disks of the TSM server to be able to have 
continguous operations.

Fault tolerant disks are to save you from angry users, whos documents 
cannot be restored, due to the fact that the TSM admin saved 300$ on disks 
and thought he was an economic genius.

Look at monolithic disks like the EMC 9000 series and the IBM ESS. You 
cant even configure RAID-0 or JBOD.

And, the save of money is to low to defend the possibility of loss of 
data.

Best Regards

Daniel Sparrman
-----------------------------------
Daniel Sparrman
Exist i Stockholm AB
Propellervägen 6B
183 62 TÄBY
Växel: 08 - 754 98 00
Mobil: 070 - 399 27 51



"Stapleton, Mark" <mark.stapleton AT BERBEE DOT COM> 
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
2004-07-21 15:39
Please respond to
"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>


To
ADSM-L AT VM.MARIST DOT EDU
cc

Subject
Re: storage pool raid 1?






From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On 
Behalf Of Daniel Sparrman
>I'd say that depends on the level of security the organisation 
>requires. 
>If the disks would crash before migration, it would mean 
>you've just lost 
>all the backup data since yesterday night. Would you like to 
>go to your 
>boss and inform him about that and explain to him that you saved 300$ 
>worth of disk?

True. However, this is an attempt to cover a double failure--the TSM
client and the TSM server disk. The only time I've seen a double failure
occur is during a complete system catastrophe--a server room fire, for
example--in which case the only redundancy that will save you is a
backup to tape that goes offsite.

>Also, if you dont migrate every morning, the loss of data 
>could be alot bigger than just one nights backup.

Also true. However, the only systems I've set up (or seen) that don't
perform daily migrations to primary tape pools are disk farms that
*never* migrate to tape; in these cases, I can certainly see RAID5
redundancy as necessary.

In my experience, I see almost no case for fault tolerance in TSM-based
disk systems that normally perform a daily migration to a tape-based
primary storage pool. If a system crashing prior to migration worries
you, perform your migration as early in the day as you can after clients
finish running their backups to a cached disk pool.

--
Mark Stapleton