RE: cluster upgrade stratedy


-----Original Message-----
From: Don.Zong@xxxxxxxxx [mailto:Don.Zong@xxxxxxxxx]
Sent: March 17, 2007 10:09 AM
To: Info-VAX@xxxxxxxxxxxx
Subject: cluster upgrade stratedy

Hi,

We are thinking upgrade our 3-nodes cluster openvms 8.2-1 itanium to
openvms 8.3. Right now every node boots off the the same system disk
which is on eva3000.

The upgrade approach we are trying to do

1. clone a system disk say disk B
2. let node A,B boot from the original system disk
3. let node C boot from disk B ( change logicals for some common files
ie uaf , so node C use the same file as node A, B does)
4. upgrade node C
5. testing on node C
6. if everything is fine, let node A, B boot from disk B
7. after a couple of months, if everything is ok remove the original
system from cluster.
8. done

Wondering if anything I've overlooked ? Any suggestions? comments?

Thanks
Don


What you are describing is essentially known as a rolling upgrade.

If it is desired to have this capability, I would recommend you maintain
a second system disk like you described and have a common disk with all
of the common files recommended in the OpenVMS doc's under recommended
cluster config's using a second system disk.

[note that whereever I say "disk". I mean a HW RAID disk]

If your applications are cluster aware, then with some basic user and
connection load balancing, you can establish an environment whereby you
can shut down systems proactively with *zero* application availability
impact - even for system OS upgrades.

Regards

Kerry Main
Senior Consultant
HP Services Canada
Voice: 613-592-4660
Fax: 613-591-4477
kerryDOTmainAThpDOTcom
(remove the DOT's and AT)

OpenVMS - the secure, multi-site OS that just works.

.

Relevant Pages