On Mon, 14 Dec 1998 [EMAIL PROTECTED] wrote:
> 2.1.131-ac11 + raid0145-19981214-2.1.131-ac9 (no autostart)
>
> 4-way P6-200, 1G ram, 10 9.1G Seagate drives (9 in raid-5 with 1 spare)
> across two aic-7880 controllers (on the MB)
>
> -------Sequential Output-------- ---Sequential Input-- --Random--
> -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
> MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU
> 2000 7141 98.0 21789 96.3 10677 79.8 9724 92.6 23544 79.5 574.4 14.4
>
> Looks great!
looks like it's pretty much CPU-bound. Time to improve the RAID5-overhead
...
does decreasing/increasing chunksize change the results? Even chunksize 4k
makes sense under certain circumstances. (in case the array is not yet
used, changing chunksize destroys data content otherwise)
-- mingo
