Hi Guys,
I have been running bonnie++ over XFS + RAID5, and I have to say
congrats on a great job. You have made my life a lot easier, and I
probably owe each of you quite a few beers.
My question though concerns large sequential reads (45K to 64K reads). I
noticed that I was not able to increase the blocksize on XFS (from 4K to
16K or 32K). Is there another way to improve these reads? Is there going
to be improvements in the kernel code or XFS in the near future so large
sequential reads are efficient???
I would like to hear a response to this, even if it says "no way are we
going to do this...".
Anyway, once again, congrats on a great job guys.
Cheers,
Kal.
|