Re: what is fsck's "slowdown"?

From: Marc G. Fournier <scrappy_at_hub.org>
Date: Sat, 4 Sep 2004 00:40:49 -0300 (ADT)
On Fri, 3 Sep 2004, Chris Laverdure wrote:

> On Fri, 2004-09-03 at 21:14, Giorgos Keramidas wrote:
>> On 2004-09-03 18:01, "Marc G. Fournier" <scrappy_at_hub.org> wrote:
>>>
>>> load: 0.99  cmd: fsck 67 [running] 15192.26u 142.30s 99% 184284k
>>> /dev/da0s1h: phase 4: cyl group 408 of 866 (47%)
>>>
>>> wouldn't it be possible, on a dual CPU system, to have group 434 and above
>>> run on one process, while group 433 and below running on the second, in
>>> parallel?  Its not like the drives are being beat up:
>>
>> My intuition says that if metadata of the first part of the disk references
>> data residing on the second part synchronization and locking would probably
>> be a bit difficult; actually very difficult.
>
> My intuition tells me that it would be a much better solution to run
> multiple fsck's concurrently. What harm could there be in fscking (num
> of processors) partitions at the same time?

I already do that ... the problem file system is the 100+GB one that just 
took 12hrs to run through on a machine that was 50% idle for the whole 
time :(  I'm looking forward to moving to 5.x because of the bkgd fsck, 
*but* ... if it takes 12hrs as a foreground process when it can suck all 
the CPU it wants, how long is it goin to take on a live server where its 
sharing with several hundred other processes? :(

----
Marc G. Fournier           Hub.Org Networking Services (http://www.hub.org)
Email: scrappy_at_hub.org           Yahoo!: yscrappy              ICQ: 7615664
Received on Sat Sep 04 2004 - 01:40:50 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:10 UTC