Enable write cache, at least for building the raid. If you want a block size of 256KB, then you sure as heck better be doing a lot of large block sequential I/O. If you are doing database work, then SQL Server reads/writes 64KB at a time, so you want to configure the disks so that you
remember a block size of X means that whenever you do a write, the RAID controller must always write X blocks at a time. Since the database writes 64KB at a time then you want to make sure that, for efficiency, you configure A RAID5, then you have a 4+1 arrangement where each disk writes a chunk of 16KB at a time. With 4 data drives + 1 parity. If you have a 6+1, then you leave some I/O on the table, so to speak. That 64KB write wont involve all of the disks. Some will be idle. (But this is OK, just not optimal).
However, with 256KB block size and a 6+1, then even if you just want to write a NTFS minimum 4KB of data, then no matter what you are going to be reading/writing well over 1MB of data for a lousy 4KB request.
That is why traditionally, you should do a RAID1 for O/S + scratch table space + log files, with minimum block size for the O/S, then go with RAID10 for performance, or RAID5 for capacity or RAID6 for availability. Read up on what block size & stripe size means on your controller manual. Lots of stuff online that explain how to set optimal block size for whatever applications you are running.
