I have to apologize again for going down under, I probably don’t have the dual talent as some of my peers at EMC; blogging and working at the same time…..
I have a good idea for a startup🙂 Run a keylogger type program to constantly run on your laptop/pc/tablet and capture important achievements, interesting topics, emails etc and automatically publish those as blog posts and tweeter feeds! who wants to pick it up? I guess WikiLeaks already did that to some extent…
Today, I wanted to share some recent SharePoint 2010 infrastructure testing done by our Shanghai team and led by Frances Hu. We call those “Solutions” as they are real SOLUTIONS and not just a couple of products proven to work on our storage arrays. and before I get to the details here’s a short checklist of what’s included in that solution:
- Server virtualization – Yes, VMware 4.1
- High availability – Sure, VMware HA cluster based
- Storage tiering – Yes, EMC CX4 (we didnt get the VNXs on time) with Flash (FAST Cache), FC (System, RecoverPoint) and SATA (everything else)
- Disaster Recovery – Oh yes, vCenter SRM using RecoverPoint/SE
- Remote BLOB storage – Of course, that’s actually part of tiering, we used the native SQL RBS FILESTREAM for that.
- Efficiency – Yes, the BLOB store was compressed using LUN compression
- Backup/Replication – Check, we used Replication Manager and Metalogix Selective Restore for item level recovery.
All that in one solution which we are going to demo at EMC World 2011 (I hope you’re coming to Vegas)…
Some of the key results/figures:
- The SharePoint farm in the solution (including the SQL server) was virtualized depicting a highly available midsize SharePoint environment
- The sustained simulated maximum user capacity was 13,080 at 10% concurrency
- Search crawl performance was improved by 91% and search response time was reduced by 27% when using EMC FAST Cache on those LUNs
- 30% of disk space savings by using EMC LUN compression features on the SharePoint server BLOB store LUNs.
- 91.2% of SQL database data file storage space was freed after enabling SQL RBS FileStream
- A full-site disaster caused only 15 minutes of downtime while all farm’s VMs failed over to DR using SRM and RecvoerPoint
- Using Replication Manager 5.3.1, it took only 6 minutes to restore a 100 GB content database from a SnapView replica
Storage Layout reasoning is based mainly on cost and somewhat controversial but I like it (mostly SATA!!!) as it proves to work!
Here’s the environment architecture
For more details you can download the whitepaper from: http://www.emc.com/collateral/software/white-papers/h8139-protection-virtualized-sharepoint-wp.pdf
If you have any questions let me know…
Until next time (which I believe would be dedicated to SharePoint DR discussion).
Happy Passover/April vacation.