Mail Archives: opendos/1997/04/23/08:30:50
On Tue, 22 Apr 1997, Alaric B. Williams wrote:
> > > Fragmentation can be solved by moving fragmented files (found when the filer
> > > notes that accessing a certain file has entailed a lot of extent seeks) into
> > > contiguous areas from time to time, a sort of background defrag that works
> > > on individual files when it feels the need.
>
> > Slooow! It seems really slow!
>
> Mark thinks all my ideas will be really slow! Peasant! Grr!!! (Only
> joking - best of friends really :-)
;-))) I'm only trying to find all the week spots (whether really existing or
just imagined) beforehand. Being a sceptic may sometimes be useful -
especially in computer stuff ;-)
> Seriously, stop and think about it. The defrag thread can run purely
> in idle time, ie it's at a REALLY LOW priority. OTOH, it sits there
OK. So it has a low priority and, as such, runs last of all the processes. Now
think about a situation when this defrag task is being pushed back to the
bottom by other ones which constantly mess with some files (say mkisofs, for
example) - that way the defragmenter is either not run at all or its work is
wasted everytime the file system is written with the huge amount of data
output from mkisofs?
> invisibly sorting the disks out - so disk access is FAST and NICE!
> What's more, it might find disk errors and things while it's at it,
> and raise suitable alerts.
I agree here - the error detection would be nice. But, IMHO, it would be hard
to make the defrag work in concert with swapfile managers and caches. I don't
know... just my opinion...
- Raw text -