Re: ext2fs "performace"

Stephen C. Tweedie (sct@dcs.ed.ac.uk)
Thu, 27 Jun 1996 11:22:52 +0100


Hi,

On Fri, 21 Jun 1996 07:59:59 +0200 (MET DST),
koenig@tat.physik.uni-tuebingen.de (Harald Koenig) said:

>> A 1 GB file on a 1k block ext2 filesystem will have 4096 indirect
>> blocks and a few dindirect blocks. Deleting the file will involve
>> essentially doing a random-access seek and read of each of these
>> blocks, so if it takes 100 seconds you are getting over 40 seeks/reads
>> per second.

> sounds reasonable (sigh;-) and matches the Bonnie result (my bonnie version
> is from 1991 if this matters):

> first I thought it's not possible ramdom seeks take all the time since
> usually I can hear such seeks on this disk quite good but removing
> this big file was almost unhearable.

The delete is not going to cause a great deal of seeking as such. It
will be accessing blocks at intervals of about one every 256KB on the
disk. Most modern drives are really very quiet when seeking just one
track at a time, and it's likely that many of the intervals are going
to be within the same cylinder anyway (in which case all the delay you
get is the rotational latency, not the head seek).

Cheers,
Stephen.

--
Stephen Tweedie <sct@dcs.ed.ac.uk>
Department of Computer Science, Edinburgh University, Scotland.