Sharepoint Forum

Ask Question   UnAnswered
Home » Forum » Sharepoint       RSS Feeds

Disaster Recovery Testing - "re-creating the farm"

  Asked By: Keaton    Date: Mar 12    Category: Sharepoint    Views: 1479

I think I'm clear on "rebuilding our server farm" from all the great
info/whitepapers/webinars/etc that I have watched so far. However, I
have yet to actually "try" this (that will be the true test). It
doesnt appear "difficult" on the surface, however it does appear
very "methodical" on what comes first, second, third, etc. (which
sometimes can be complex).

Anyhow, these particular questions are more from a "high-level"..

Here is my problem:
We have a medium server farm in production (2 FEs, 1 job/index, 2-
node A/P SQL Cluster). In our SPS2k3 environment, we have 12 virtual
servers each with a separate portal. In addition, we have a 'root'
portal called "corporate" that is the "Shared Services Home". The
other 12 portals (Ie. ITS, Finance, Manufacturing, HR, etc.) are
consumers of the Shared Services from "corporate". All run neatly
under one application pool (so far). We have approx 5,000 users.

In building out/demonstrating our Disaster Recovery plan, my boss has
asked me to "prove" to the business that I can restore a portal or
restore the entire farm successfully.

Questions I have:

- Can I actually "rebuild" our SPS2k3 farm on separate hardware while
the "original"/"production" farm is still online and in production?
- I do have VMWARE and a nice big box, to "virtually" re-create all of
these servers on or off the network.
- Would I be better off "rebuilding" this environment "off the
network" or on a "private/offline" network? I assume I would need to
miminc all my Windows Infrastructure services as well? (Ie. AD, DNS,
WINS, etc) or no?
- What is my quickest/easiest path to "proving" that we can recover
from a "disaster" with our current backups/solutions/procedures?

More about our backups:
- we use Veritas NetBackup Adv Client w/ SharePoint Agent (backs up
local filesystem, system state, metabase) and the SPSAgent backs up
the portal databases and the index (much like SPBackup.exe)
- In addition, I wrote an application to automatically backup our
entire portal/index/etc. using the native SPSBackup.exe utility.
- We are also using SQL Enterprise Manager to backup the databases.

The combination of the three may seem redundant, but I feel safer with
multiple backups to different media (Ie. disk, tape, etc) to speed
recovery should a disaster strike. If I have to use NTBackup just to
prove out this "DR" experiment I have been tasked with that is fine.
I will do whatever it takes to "prove" that this can be done and ease
the worried minds.

Question Summary:

- How can I re-create my production environment in a lab to prove that
it can be done? and what (from a high-level) would this scenario look
like? (Ie. Virtually,Infra, Network, on production network, offline,



No Answers Found. Be the First, To Post Answer.

Didn't find what you were looking for? Find more on Disaster Recovery Testing - "re-creating the farm" Or get search suggestion and latest updates.