From: marcf@nexus.yorku.ca (Marc G Fournier) Subject: Re: Disk Quotas (was Re: New feature for the filesystems.) Date: Thu, 15 Apr 1993 13:12:33 GMT
davidsen@ariel.crd.GE.COM (william E Davidsen) writes:
>In article <s_fuller.734838052@vincent2.iastate.edu>, s_fuller@iastate.edu (Steve Fuller) writes:
>| 1) If the file is in the middle of a write when you exceed your
>| quota, the file will be written in its entirety and not
>| truncated. Any furthur write requests will be denied
> I certainly don't want a quota like that. If the hard limit isn't hard
>it doesn't protect the system. unless writes STOP at the hard limit the
>user could run the system out of disk anyway.
but, isn't that what the ulimit is for? to control the size of
the file that a user creates? if that were the case, then why not
hhave the 'hard limit' be that a user cannot create a file, period,
if they have reached it, and have the ulimit being there to keep them
from just continually running. Most ulimit's I've seen are defaulted
to about 2Meg or so. Understandably, if their hard limit is 1Meg,
then creating a 2Meg file starting at their .5Meg level would allow
them to pass their hard limit by 1.5Meg, but in that case, why would
their ulimit be larger then their hard limit?
just a thought...
marc