On Wed, Dec 20, 2000 at 12:56:31AM +1300, Godfrey Livingstone wrote:
> The raid 0.9 does readbalancing but Mika's patch refines reads further.
interesting patch. Basically it takes into account the last position of each
disk before choosing which disk to read from. But I've some remark that I'd
like to see somehow addressed before inclusion:
1) it doesn't take into account seeks generated by writes so
it seems to me it could take the worst decision sometime
2) it thinks backwards seeks are as expensive as forward seeks
3) this fast path check is wrong, if the check returns true it means
the disk will do a seek backwards while there could be another disk
that could read the block without any seek
+ if( this_sector == conf->mirrors[new_disk].current_position )
+ goto rb_out;
Andrea
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to [EMAIL PROTECTED]
