Patrick Welche <prlw1%cam.ac.uk@localhost> writes: > START layout > 64 1 1 5 > > That means that I can write 64 x 512 x 4 bytes = 128k in one go? (Block > size always 512 bytes or real physical disk block size assuming disk > isn't pretending?) > > If I create a wedge of type raidframe on each of the four disks, the > offset should be 64? 256? (either way not the default 34...) There are two separate issues. The one everyone worries about is how a logical write of a reasonable size to /dev/raid ends up being handled. Most of the documentation addresses this, and it's generally about 64K chunks. The other is how sectors line up on the disk. For older disks with 512-byte physical sectors, it probably doesn't matter. But 34 is a goofy number in the modern age, and it's not a multiple of 8 (for 4K physical sectors). 64 seems like a good value. I believe the RF header is 64 (512-byte) sectors. The key point is that 32K is a multiple of 4K, for the newer disks with 4K physical sectors. So I would create wedges at 64. But I also would not use RAID5 in software. I just use RAID1, which seems simpler and avoids the RMW issue for writes less than a stripe.
Attachment:
pgpbAb6bYcIuq.pgp
Description: PGP signature