And on this farm they had a 2007 upgrade that needed to happen. After a Microsoft Consulting Services review, they were told to fix up a number of issues in their current environment:
- Get web/app/DB servers in the right place.
- Intranet and internet portals on shared disk spindles, they want to move to same hardware (SAN) w/dedicated spindles.
- One content DB was 210 GB. That DB needs to lose some weight.
- They had 40,000 resources in a single resource group… Microsoft said they needed to change this (db size and resource #’s) because they were outside supportable boundaries. Basically, this becomes a real bear on the system, and it’s really difficult to backup that many files.
They are using a DMX3 Tier1 146GB 10K for content DB. A DMX4 is being purchased for this environment and they will share this will other applications in the company. No problems with the storage.
For backup they use the point product AvePoint for MOSS backup. They use a multi-phase process to push Dev to Test to Cert to Production. Seemed OK, but then they thought a second and said.
“You know what, there is an opportunity for a better backup product to push data through the dev to production cycle.”
I asked more about this and found, like many customers, they deploy MOSS farms in stages. The current stages of their MOSS lifecycle were:
1. Dev – Virtual everything (DB, Web, App)
2. Test – Virtual everything (DB, Web, App)
3. Cert – Virtual web, app, physical DB (same as prod)\
4. Prod – Virtual web, app, physical DB
Since EMC is known for it’s proficiency in virtualization and data protection, they thought we might have a slick product (like Replication Manager or Networker) that could assist in the movement of data along these physical and virtual tiers without having to backup and restore every time. Well, yes we do.
Well, I told them, it’s great timing. The Networker team has not only made the product much simpler to use – the Fast Start edition of the product can be installed in under 20 minutes – they also have been working really hard on their SharePoint support. The new Networker Module for Microsoft Applications (NMM) they offer the trifecta of Sharepoint backup and restore functionality: network backups, VSS application consistent backups, and fine grained item-level restores. Whitepaper here (or click the picture below).
Next step: the customer is going to test this with EMC’s Proven Solutions group, where EMC works with Microsoft to demonstrate our technologies working together. For SharePoint, we test large and small farms in physical and virtual server environments, load them up, and see where they break. The people doing the testing are having a blast doing cool stuff every day to push the limits of datacenter technology. Sometimes the work they do is more Mythbusters (isn’t performance worse on virtual? BUSTED), and sometimes it’s more like CSI, which usually starts with a dead body and then a lot of time is spent trying to see what the hell happened :)
I do know each day is different and fun for them, so the resulting documents you see are extremely good. Unfortunately about half of it is buried behind EMC’s internet portal PowerLink, but that is being fixed…
Here’s a sample of what is available publicly at the moment:
- Saving 74% of your Power Bill with SharePoint Virtualization
- Top Ten Reasons for EMC with SharePoint
- Best Practices: SharePoint on Celerra
- Solution Sheet: SharePoint Backup with Networker and Hyper-V
- Reference Architecture: SharePoint07, SQL05 and ESX3.5
- EMC Consulting Services: Portals and Collaboration