Solaris Volume Manager (SVM)
RAID0 is striping, in which data is spread across multiple spindles (disks). Data written to the RAID0 is broken up into chunks of a specified size (interlace value) and spread across the disks.
For instance, if you put 8192 interlace, one 8K page can involve just one spindle. If you read 65536, and have 8 disks, you read 8 pages at once, subject to seek time, rotational latency and conflicts for bus or controller.
If your disk average spin covers always just less than 262144 bytes, a big interlace like that would mean that for one seek, you pick up most of the track, and a good design can start in the middle, killing rotational latency.
It is a tuning decision. Certainly, should be some multiple of the size of the disk record. Extra data is often fetched by buffering or caching to RAM somewhere along the read chain. UNIX data is not always laid out sequentially on a disk.
Good luck!
