Today I tried to do some rawhide builds in my usual mock setup (running
on updated x86_64 F11), but somehow I can't even init a chroot due to:
Executing command: ['/usr/sbin/groupadd', '-g', '498', 'mockbuild']
Child returncode was: 4
GID 498 is already occupied by exim, which gets pulled in beca
On Thu, 2009-07-30 at 21:44 -0500, Jason L Tibbitts III wrote:
> Today I tried to do some rawhide builds in my usual mock setup (running
> on updated x86_64 F11), but somehow I can't even init a chroot due to:
>
> Executing command: ['/usr/sbin/groupadd', '-g', '498', 'mockbuild']
> Child returnco
> "JK" == Jesse Keating writes:
JK> Mock is trying to add a user / group that matches the user / group
JK> calling it. Is the user calling it of gid 498?
That's the GID of the mock group on the host. It's not my primary GID,
but I guess that doesn't matter. It would explain why I'm the on
On Thu, 2009-07-30 at 23:02 -0500, Jason L Tibbitts III wrote:
> That might help this specific case, I guess, but perhaps there's a more
> general problem. If mock absolutely requires that it be able to create
> a group in the chroot with the same GID as the mock group on the host,
> then perhaps
> "JK" == Jesse Keating writes:
JK> Hrm, so I wonder about this. Does exim rely on the group ownership
JK> at all for anything? Would it make sense to have a general
JK> 'service' or 'nobody' group that these things could be tossed in if
JK> the group isn't to be used, to avoid taking dynam
> "JLT" == Jason L Tibbitts writes:
JLT> So useradd must have changed its behavior quite recently.
It could be shadow-4.1.4.1-sysacc.patch, I guess, but that was built in
rawhide on the 16th of this month and I've done plenty of builds since
then.
- J<
--
fedora-devel-list mailing list
f
On Fri, 2009-07-31 at 10:41 -0500, Jason L Tibbitts III wrote:
> I don't think it's bad for exim to groupadd 93 first, but honestly I
> don't know what happens to existing installations that may have a
> different GID set up and I don't want to break anything. I guess such
> systems would be runni
Jason L Tibbitts III wrote, at 08/01/2009 12:46 AM +9:00:
"JLT" == Jason L Tibbitts writes:
JLT> So useradd must have changed its behavior quite recently.
It could be shadow-4.1.4.1-sysacc.patch, I guess, but that was built in
rawhide on the 16th of this month and I've done plenty of builds s