[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: quotas



On Tue, Mar 28, 2000 at 04:55:13AM +0000, Jim Breton wrote:
> Hello, here is a message I sent to the linux-admin list.  I am
> re-sending here because I have another question following this one.
> 
> ************
> Is this normal and just undocumented?  Or is it documented somewhere...
> or is it a bug?
> 
> I was having some trouble getting quotas to work and after some fiddling
> I discovered the following behavior.
> 
> If I have the following line in my /etc/fstab everything is fine:
> 
> /dev/hda2   /    ext2   defaults,errors=remount-ro,usrquota    0     1
> 
> 
> But previously I had been using this:
> 
> /dev/hda2   /    ext2   defaults,usrquota,errors=remount-ro    0     1
> 
> 
> The lines are the same except for the order of the last two mount
> options.
> 
> Well, if I use the second method, instead of creating "/quota.user" and
> storing quota information in that file, quotacheck writes the file into
> "remount-ro" (a file with that name) into my current directory!

bug #46610 its fixed in the current potato quota.  (or so the BTS
says)

another solution is to set the error behaviour in the superblock
and remove the mount option, i did this because for some reason the
`errors=remount,ro' was showing up twice in mount, which i thought was
ugly ;-)

drop down to single user mode, umount all filesystems, remount /
readonly, run:

tune2fs -e remount-ro /dev/hda? <-- your root partition.

your probably better off rebooting at this point.  you can then remove
the errors=remount-ro from your fstab file. 

> 
> OK it does seem that the quota utils are thinking the "=remount-ro" is
> an argument, and they should not be doing this.  It also does this on a
> Red Hat 6.0 system I tried (as well as the Slack 7 as I've mentioned).
> Should this be reported as a bug to the author/maintainer?

already done, already fixed.  quota 2.0 is alot better from what i
have heard, but not done yet and as a result not in potato.

> Also, I have been having a problem implementing quotas on my slink
> machine.  I am running a 2.2.14 kernel which I think may have something
> to do with it.
> 
> I have an account "jim" which "repquota" claimed has about 1500000
> blocks in use.  While I was testing out the quotas, I set jim's soft and
> hard limits to 2000000.  So, he should have been fine -- but for some
> reason repquota says he is over quota!

not sure about this one... check the bugs on quota at:

http://cgi.debian.org/cgi-bin/pkgreport.cgi?archive=no&pkg=quota

there are a few that sound similer to your situation but not quite,
see for yourself...

-- 
Ethan Benson
http://www.alaska.net/~erbenson/

Attachment: pgpQrogfVYAPu.pgp
Description: PGP signature


Reply to: