- Yet to hear about clustering by itself being source of Store corruption!
- Properly set-up clusters significantly boost Exchange uptime and
capabilities for performing maintenance tasks without impacting users - in
the middle of the day if you will. (With Exchange Server 2003 on server side
and Outlook 2003 Cached Mode on client side, most users never realize a
failover occurred).
- Windows Server 2003 = better SAN support, including iSCSI
- Windows Server 2003 + Exchange Server 2003 = major improvements in
clustering that have probably increased number of cluster deployments
significantly compared to 2000, and certainly compared to 5.5, imo.
http://www.microsoft.com/technet/prodtechnol/exchange/guides/WhatNewE2k3/feced74
a-2eb3-40a7-842f-f7129b2ab3cc.mspx
- Store HAS changed from 5.5 to 2003, has many improvements, some listed
here:
http://www.microsoft.com/technet/prodtechnol/exchange/guides/WhatNewE2k3/e14c602
8-9f8d-41a7-b981-4b1e7e867b3a.mspx
- Sidenote: Store on shared disks is still a single point of failure unless
replicated. Solutions available from 3rd parties (NSI DoubleTake, EMC,
NetApp SnapMirror, etc.). Exchange 12 is reported to include log shipping.

Signature
Bharat Suneja
MCSE, MCT
www.zenprise.com
blog: www.suneja.com/blog
-----------------------------------
> Currently our environment consists of Windows 2000 clusters running
> Exchange 5.5, these clusters are SAN attached. We have experienced
[quoted text clipped - 19 lines]
> regards
> P