Upgrading lots of TSM Clients and changing the TSM Server

w8er

Active Newcomer
Joined
Mar 22, 2010
Messages
11
Reaction score
0
Points
0
Hi Everyone,

I am pretty new to this forum and very new to TSM. My job is to upgrade our "TSM" farm from 5.5 to 6.x. (Consisting of 4 Servers)

Starting with the first server running on an old hardware and with an old name not meeting the new naming conventions. So it is going to be the New Server Network" method i guess.

Is there anything out there which makes the client configuration easier or unattended ? I need to change the TSM Server Name on about 100 tsm client servers and upgrade clients to the latest version.

many thanks in advance for any tips and tricks
dave
 
Why the push to 6.X? It's still plagued with problems! I would move to 5.5 and wait.

As for mass client upgrades, we do one at a time using an install script. Basically log on run the script and move to the next one. If you are good at multitasking you can have three to four upgrades running at one time. When directed to upgrade (not often) I'll dedicate about 1 hour each day and do 20-30 clients each day.

That being said we still have some clients running 5.3.4.2 client code. They work just fine. No reason to upgrade unless there is a security risk or fix that you require.

BTW... are you chaning the actual server name or just the TSM node name?
 
Hi Jeff,

thanks for your answer. we want to test the client deduplicaton feature since we are backing up remote sites over the wan. therefore goto v6.2.

we have some respect of the DB2 database.

we would change the server name and keep the nodename.

what are the major bugs in 6.2 ?

thanks
david
 
Well.... I'm not sure how conservative your company is but my organization would frown upon using a code level that was released in the last couple of days. Is the dedup worth the risk of deploying brand new code?

We too are backing up our remotes via commercial wan. We have enabled client side compression which is good for about a 45% crunch over all. I’m not sure how much we would benefit from source side dedup unless it was a file server with multiple copies or similarly daily changing files.

If you do deploy it please start a thread with your findings.
 
Back
Top