stephen> On Mon, 13 Sep 2004, Daniel Feenberg wrote: >> On Mon, 13 Sep 2004, Michael C Tiernan wrote: >> >> > On Sun September 12 2004 17:45, Dean Anderson wrote: >> > > It depends on what you mean by 'monitoring'. >> > Thank you for replying (everyone). >> > >> > I'm being asked[1] for: >> ... >> > 2) An X-Y graph of part of a file system, preferably, one graph >> > for each user (10+/-) space showing use and change. >> > Realistically, this should be done (for some, and thereby >> > preferably all, of the users) once an hour. This is the hard >> > one. Some of the users have SO much space that a 'du' (we >> >> "quota -v" is what you want. Just for curiosities sake, I did: >> >> cd /home;quota -v * >/tmp/file.foo
stephen> close, but a better way might be to use quot(8) (or quot(1M), stephen> depending on the system): Quot doesn't work (at least on my solaris boxes) on anything but UFS filesystems. What he really wants is to be able to get emailed an hourly quota report from the filer (god knows why you care about space on an hourly basis) which can be parsed and then filtered as need be. If the admins of the NetApp won't work with him, then I'd just threaten to run multiple DU commands from multiple clients all at the same time to bring the box to it's knees. Then when that admins complain, point them to the boss and say "he/she wanted it like this, talk to them". And then sit back and watch the merry discussion. I'd bring popcorn. Basically, you should already have quotas on your trouble users, and they can manage their disk space themselves. When they can't write any more data, they're out of space. Presto, they then get to cleanup. John _______________________________________________ bblisa mailing list [EMAIL PROTECTED] http://www.bblisa.org/mailman/listinfo/bblisa
