Comments on ESIX's Fast File System

[Glenn Geers]

Hi,
	The ESIX (rev D) fast file system appears to be quite odd in
	it's behaviour.  I have a 30 Mb filesystem with 16 Mb free; the
free space is composed of 2500 free *fragments* and the rest in blocks.
Fragmentation is at 7.5%. From this it appears that the ESIX
implementaion is breaking up blocks rather than looking for free
fragments when it writes a file that is not a multiple of the blocksize
(8 kb in my case; can be configured to 4 kb). Also, the minfree
parameter is not adhered to. True BSD ffs implementations allow only
90% of the available space to be used by ordinary users. On ESIX 100%
of the available space can be written to by anyone.

Anyway, I guess this is their first release of FFS so we should be
patient.  But please, if anyone from ESIX is listening, provide a fix
for these little problems and while your at it, provide symbolic links
and long filenames so we can be V.4 ready.

				Just my $0.02 worth,
							Glenn

glenn at qed.physics.su.oz.au


--
Glenn Geers                       | "So when it's over, we're back to people.
Department of Theoretical Physics |  Just to prove that human touch can have
The University of Sydney          |  no equal."
Sydney NSW 2006 Australia         |  - Basia Trzetrzelewska, 'Prime Time TV'