Hello,
On Sun, Oct 01, 2000 at 12:45:47PM -0700, LA Walsh wrote:
> Forgive me if this has been asked before, but has there ever been any
> thought of having a 'nice' value for disk accesses?. I was on a
> server with 4 CPU's but only 2 SCSI disks. Many times I'll see 4 processes
> on disk wait, 3 of them at a cpu-nice of 19 while the foreground processes
> get bogged down by the lower priority processes due to disk contention.
I'm considering how disk "bandwidth" use by users may be controlled, and some
coding has been started. But it's not what may be announced as a
break-through yet.
The main problems are:
- what to consider as the resource?
(I'm inclined to account for the time the requests of the given "user" are
served by the disk)
- how to ensure an acceptable level of fairness in the case of considerable
amount of "users" (more precisely, subjects of accounting)?
I think, the question is not about ensuring bounds on latency, and the
request order should be determined by dynamic priorities of users in a
large degree.
Best regards
Andrey V.
Savochkin
-
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/
This archive was generated by hypermail 2b29 : Sat Oct 07 2000 - 21:00:11 EST