A section in the TechNet article Best Practices for Operational Excellence (SharePoint Server 2010) states the following:

Separate system components into logical drives and use RAID for redundancy.
Components on drive                             Recommended RAID level
Windows and program files drive                 RAID 1
Operating system swap drive and temp directory  RAID 1
Log files                                       RAID 1
Boot disk for imaging and                       RAID 1
  Windows Desktop Search (optional)
Use at least four physical disks and use separate disks to keep the log files and swap drive separate from the Windows and program files drive.

If I wanted to build one server according to this advice, what would it actually look like? When it says "use at least 4 disks" for RAID, then am I looking at getting 16 disks? Or is this saying the underlying physical structure should be 4 disks, and I should create logical drives on top of that RAID structure?

Lastly, how would this advice translate to a virtual environment?

1 Answer 1


Your performance and redundancy requirements translate into an appropriate disk configuration.

e.g. If it's a really busy server and you require maximum redundancy, you might choose to separate out file types logically into RAID 10 volumes.

In my fairly limited experience this is normally considered overkill except in the case of SQL server. In that case, you could for example have:

  1. 2 * disks in a RAID 1 configuration (OS/system)
  2. 4 * disks in a RAID 10 configuration (TempDB)
  3. 4 * disks in a RAID 10 configuration (SQL data)
  4. 4 * disks in a RAID 10 configuration (SQL logs)
  5. ...additional volumes as required, e.g. for search databases/logs

For Web/app servers my standard configuration is:

  1. 2 * disks in a RAID 1 configuration (OS/system)
  2. 4 * disks in a RAID 10 configuration (everything else)
  3. ...additional volumes as required, e.g. search indexes

As for virtual environments, I've found that using VMs add a layer of abstraction that can mask underlying storage issues (it can be difficult enough with physical servers). Most MS documentation suggests utilising pass through disks for direct storage access as opposed to storing data inside VHD files (especially for SQL server).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.