carl@probitas.cs.utas.edu.au (Carl Lewis) (11/22/90)
Ok I give up, does Aix (3.1) have quotas. Some routines return error
on exceeding quota, but I really get the feeling this is a hangover
from older documentation .
Does there exist a quota system , and if so what marvellous acronym,
psuedonym or just incorrect labelling have IBM lumbered it with ?
Thanks a lot,
Carl.
--
Snail: Computer Science Dept., University of Tasmania,
GPO Box 252C, Hobart, TAS, 7001, Australia.
AARN : carl@tasis.eecs.utas.edu.aujerry@heyman.austin.ibm.com (Jerry Heyman) (11/24/90)
In article <carl.659244255@probitas> carl@probitas.cs.utas.edu.au (Carl Lewis) writes: >Ok I give up, does Aix (3.1) have quotas. Some routines return error >on exceeding quota, but I really get the feeling this is a hangover >from older documentation . > AIX v3.1 DOES NOT support the quota from Berkley. I have not seen the routines that report that quota has been exceeded, but I would be curious in finding out what routines do this. >Thanks a lot, >Carl. Your welcome, jerry heyman > >-- > >Snail: Computer Science Dept., University of Tasmania, > GPO Box 252C, Hobart, TAS, 7001, Australia. >AARN : carl@tasis.eecs.utas.edu.au -- Jerry Heyman IBM T-R: jerry@heyman.austin.ibm.com AWD Tools Development VNET : HEYMAN at AUSVMQ AWD Austin T/L : 793-3962 *** All opinions expressed are exactly that - my opinions and NOT IBM's
carl@probitas.cs.utas.edu.au (Carl Lewis) (11/26/90)
In <4317@awdprime.UUCP> jerry@heyman.austin.ibm.com (Jerry Heyman) writes: >In article <carl.659244255@probitas> carl@probitas.cs.utas.edu.au (Carl Lewis) writes: >>Ok I give up, does Aix (3.1) have quotas. Some routines return error >>on exceeding quota, but I really get the feeling this is a hangover >>from older documentation . >> > >AIX v3.1 DOES NOT support the quota from Berkley. I have not seen the routines >that report that quota has been exceeded, but I would be curious in finding out >what routines do this. The open, openx, creat, rename, link, mkdir, symlink, mknod, mkfifo routines all come up under an info search for quota. All the ones I've looked at return EDQUOT on exceeding quota !!!. Since you've confirmed our suspicions that quotas do not exist it would be interesting to find out where the EDQUOT hangover is from :-) Replying with a post, 'cos as lots of people have noted, the mail still bounces! > >>Thanks a lot, >>Carl. > >Your welcome, > >jerry heyman >> >-- >Jerry Heyman IBM T-R: jerry@heyman.austin.ibm.com >AWD Tools Development VNET : HEYMAN at AUSVMQ >AWD Austin T/L : 793-3962 >*** All opinions expressed are exactly that - my opinions and NOT IBM's -- Snail: Computer Science Dept., University of Tasmania, GPO Box 252C, Hobart, TAS, 7001, Australia. AARN : carl@tasis.eecs.utas.edu.au
jerry@heyman.austin.ibm.com (Jerry Heyman) (11/27/90)
In article <carl.659576268@probitas> carl@probitas.cs.utas.edu.au (Carl Lewis) writes: [... discussion of whether or not quota command exists where I asked for examples of where this was being returned ...] > >The open, openx, creat, rename, link, mkdir, symlink, mknod, mkfifo >routines all come up under an info search for quota. All the ones I've >looked at return EDQUOT on exceeding quota !!!. Since you've >confirmed our suspicions that quotas do not exist it would be >interesting to find out where the EDQUOT hangover is from :-) > Since I'm unfamiliar with the interanlas of any of the subroutines that you mention, I looked up 'mkdir' in info. While its true that EDQUOT can be an error from that call, you will see that it is returned for one of two reasons: the first would be for exceeding disk quota (which AIX v3.1 does not support); the second is if i-nodes become exhausted for a particular filesystem. The only other possible reason would be for compatability for code ported from BSD that expects that as a possible return. >Replying with a post, 'cos as lots of people have noted, the mail still bounces! > I know, and we're slowly trying to resolve that. :-) jerry -- Jerry Heyman IBM T-R: jerry@heyman.austin.ibm.com AWD Tools Development VNET : HEYMAN at AUSVMQ AWD Austin T/L : 793-3962 *** All opinions expressed are exactly that - my opinions and NOT IBM's
jfh@rpp386.cactus.org (John F. Haugh II) (11/27/90)
In article <4317@awdprime.UUCP> jerry@heyman.austin.ibm.com (Jerry Heyman) writes: >In article <carl.659244255@probitas> carl@probitas.cs.utas.edu.au (Carl Lewis) writes: >>Ok I give up, does Aix (3.1) have quotas. Some routines return error >>on exceeding quota, but I really get the feeling this is a hangover >>from older documentation . > >AIX v3.1 DOES NOT support the quota from Berkley. I have not seen the routines >that report that quota has been exceeded, but I would be curious in finding out >what routines do this. OK, this is the real meal deal. The security department was responsible for porting/implementing the BSD-style quota mechanisms. For reasons not worth going into, that functionality was not put in AIX v3.1, but is/was scheduled to appear in a "future" release, whatever that means (I don't speak for IBM, that's one thing it means ;-). You will =never= get EQUOTA or whatever returned from a system call because you ran out of file system space or any other resource. I've never tried to run out of CPU time, but I seem to recall that the CPU time limit is also not enforced. As others have pointed out, there are a number of attributes in /etc/security/limits which are documented as not being used. Yup, they aren't used. -- John F. Haugh II UUCP: ...!cs.utexas.edu!rpp386!jfh Ma Bell: (512) 832-8832 Domain: jfh@rpp386.cactus.org "SCCS, the source motel! Programs check in and never check out!" -- Ken Thompson
guy@auspex.auspex.com (Guy Harris) (12/03/90)
>While its true that EDQUOT can be an error from that call, you will >see that it is returned for one of two reasons: the first would be >for exceeding disk quota (which AIX v3.1 does not support); >the second is if i-nodes become exhausted for a particular filesystem. So fix it to return ENOSPC if inodes become exhausted, instead of EDQUOT, both for compatibility with the 4.3BSD file system code and the System V file system code (both of which return ENOSPC if you run out of inodes, which means that behavior probably dates back to time immemorial), and to prevent people from getting confused into thinking AIX 3.x supports quotas just because it incorrectly returns EDQUOT when you run out of inodes (rather than running out of your inode *quota* - the BSD file system returns EDQUOT when that happens).