Messages in this thread |  | | Date | Mon, 18 Sep 2000 01:38:47 +0200 | From | Andrea Arcangeli <> | Subject | Re: An elevator algorithm (patch) |
| |
On Sun, Sep 17, 2000 at 10:53:17PM +0200, Peter Osterlund wrote: > The test2 elevator (assuming it is the same as the test8 version) in the
Yes it's the same in such respect.
> infinite latency case will always send the request with the lowest sector > number to the drive. (The request queue will always be sorted, since the
Correct. And infact that's what is happening to 2.4.0-test8.
> elevator function degenerates to insertion sort.) Do you really suggest > that this is as good as a real elevator algorithm?
No, I wasn't suggesting anything like that, I was only saying "don't complain the elevator if the number you got out of test8 aren't good and you were running on scsi". I guess the latency defaults are been set to a too high default value exactly to be able to say that :). Jens also verified that a default of 250 and 500 is good enough to not degenerate the tiotest numbers (tested on 2.2.17 + elv-simple-1 patch), and with a 250/500 pair a `find` with a `cp /dev/zero .` in background is not too bad (not like an deadlock and you can visibly see the fix in action, however you'll definitely notice the cp in background :).
Andrea - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
|  |