Looks good to me from a division of the SQL file view. The main thing to keep in mind with SQL is the Disk IO. If you expect high IO to disk you need to ensure that you have enough spindles to meet the IO that the application will put on SQL
You need to estimate the balance between reads and writes to disk to decide on the RAID config needed to match the IO.
normal 10K disk can provide 120 IOPS where a 15K disk can provide 180 IOPS.
As a general rule of thumb for SQL the disks should be in a RAID 10 config if you can afford it and if your storage has the capacity
I also recommend that you dedicate disk for each of the above drive letters to ensure that you don't have any sharing of disk IO
Example 6 disks all in a single raid 5 and 5 partitions on that disk is bad for IO as the IO on the disks is shared and not so good if you need high IO on even one drive letter because each time the other disks are accessed the one that needs high IO will have delays reading or writing
If you have 4 disk in a raid dedicated to one drive letter (one raid and only one partition on that raid) this is will dedicate the IO to that drive letter and to the application (SQL) that will be accessing that disk giving best performance
There are too many scenarios to go into on RAID and the pros and cons of each but it should be in your mind when sizing for apps like SQL which can be heavy on disk IO especially on EV when you are doing expiry or also on EV when you are doing a lot of searches on the indexes as in with DA
