lkml.org 
[lkml]   [1998]   [Nov]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: elevator algorithm bug in ll_rw_blk.c
Hi,

On Sun, 15 Nov 1998 14:13:31 -0500, Chip Salzenberg
<chip@perlsupport.com> said:

> According to Alan Cox:
>> Chip:
>> > But that's a *one-way* elevator. Ideal elevators are two-way,
>> > aren't they?
>>
>> If you go for the shortest movement alone you starve the edges of the disks.

> I said "two-way", not "shortest movement".

Two-way elevators _do_ have this problem. A block in the middle
cylinder of the disk gets serviced twice per full two-way pass, once in
each direction. Maximum wait time is half a pass; average is a
quarter. For a block at the edge of the disk, it will get serviced only
when the two-way pass approaches close to the edge of the disk, and will
not get another chance for a full pass, average wait half a pass.

It's not a total stavation issue, but it _does_ give you a significant
unfairness.

--Stephen


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:45    [W:0.082 / U:0.620 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site