I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
Digging through root.log shows this message buried in the middle of all the 'yum install' output: error: can't create transaction lock on /var/lib/mock/fedora-4-i386- core/root/var/lock/rpm/transaction
So far, I've had similar results building for FC3/FC4/EL4, so I'm guessing its a host machine issue rather than a chroot issue. Has anyone else come across this already? I am using an older version of mock - 0.4-2 on one machine, and 0.4-5 on another; has something changed since then to fix this issue? I guess I need to upgrade eventually to get the newer mock config setup, etc. but I've been putting that off...
Thanks, Jeff
On Sep 8, 2006, at 5:29 PM, Jeff Sheltren wrote:
I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
Digging through root.log shows this message buried in the middle of all the 'yum install' output: error: can't create transaction lock on /var/lib/mock/fedora-4-i386- core/root/var/lock/rpm/transaction
So far, I've had similar results building for FC3/FC4/EL4, so I'm guessing its a host machine issue rather than a chroot issue. Has anyone else come across this already? I am using an older version of mock - 0.4-2 on one machine, and 0.4-5 on another; has something changed since then to fix this issue? I guess I need to upgrade eventually to get the newer mock config setup, etc. but I've been putting that off...
Thanks, Jeff
Sorry, but I meant to add - inside the chroot, the directory /var exists, but /var/lock is not getting created.
-Jeff
Jeff Sheltren wrote:
I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
patch attached, for mock 0.4, easy to adapt to 0.6, I've not looked at 0.6+ as yet, but should work there too, its trivial.
HTH
On Sep 8, 2006, at 6:06 PM, Karanbir Singh wrote:
Jeff Sheltren wrote:
I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
patch attached, for mock 0.4, easy to adapt to 0.6, I've not looked at 0.6+ as yet, but should work there too, its trivial.
Thanks, Karanbir. That was my first thought (have mock create the directory), but I wasn't sure what the cause was, so I was reluctant to fix something I wasn't sure why it broke. Was this a change to the filesystem or rpm RPM(s) or what?
-Jeff
Jeff Sheltren wrote:
On Sep 8, 2006, at 6:06 PM, Karanbir Singh wrote:
Jeff Sheltren wrote:
I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
patch attached, for mock 0.4, easy to adapt to 0.6, I've not looked at 0.6+ as yet, but should work there too, its trivial.
Thanks, Karanbir. That was my first thought (have mock create the directory), but I wasn't sure what the cause was, so I was reluctant to fix something I wasn't sure why it broke. Was this a change to the filesystem or rpm RPM(s) or what?
basically, rpm issue ( feature ? ) lots of info here : https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=151255
On Sep 8, 2006, at 9:23 PM, Karanbir Singh wrote:
Jeff Sheltren wrote:
Was this a change to the filesystem or rpm RPM(s) or what?
basically, rpm issue ( feature ? ) lots of info here : https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=151255
Cool, thanks for the link! So, can somebody verify this bug w/ the latest version of mock? If so, I think we should apply Karanbir's patch...
-Jeff
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Jeff Sheltren wrote:
On Sep 8, 2006, at 9:23 PM, Karanbir Singh wrote:
Jeff Sheltren wrote:
Was this a change to the filesystem or rpm RPM(s) or what?
basically, rpm issue ( feature ? ) lots of info here : https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=151255
Cool, thanks for the link! So, can somebody verify this bug w/ the latest version of mock? If so, I think we should apply Karanbir's patch...
This one looks like a no-brainer, but I'm doing a bit of testing just to make sure (I've been bitten by "no-brainers" before :)). Once I verify that I haven't broken anything obvious, I'll push a new tarball up for Jesse to grab and stuff into Extras.
Clark
On Sep 11, 2006, at 1:55 PM, Clark Williams wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Jeff Sheltren wrote:
On Sep 8, 2006, at 9:23 PM, Karanbir Singh wrote:
Jeff Sheltren wrote:
Was this a change to the filesystem or rpm RPM(s) or what?
basically, rpm issue ( feature ? ) lots of info here : https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=151255
Cool, thanks for the link! So, can somebody verify this bug w/ the latest version of mock? If so, I think we should apply Karanbir's patch...
This one looks like a no-brainer, but I'm doing a bit of testing just to make sure (I've been bitten by "no-brainers" before :)). Once I verify that I haven't broken anything obvious, I'll push a new tarball up for Jesse to grab and stuff into Extras.
Clark
Cool, thanks. I wasn't trying to imply that someone else needs to do the patching and testing; I'm glad to do so considering I'm the one that brought up the problem ;) But if you already have a system in place for getting updates pushed to FE I don't want to step on your toes!
Thanks, Jeff
Karanbir Singh wrote:
Jeff Sheltren wrote:
I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
patch attached, for mock 0.4, easy to adapt to 0.6, I've not looked at 0.6+ as yet, but should work there too, its trivial.
Patch works with mock-0.6.2 fine here. (:
-- Rex
Thanks, Rex. I'm also happy with it on my mock 0.4 installs, and haven't seen any ill effects :)
So, any objections to me committing the attached patch? It is simply Karanbir's patch adapted for current CVS.
Thanks, Jeff
Rex Dieter wrote:
Karanbir Singh wrote:
Jeff Sheltren wrote:
I'm not sure if it was the move to CentOS 4.4 or what, but now my mock builds are failing on CentOS 4.4 hosts where they were working fine a couple weeks ago...
patch attached, for mock 0.4, easy to adapt to 0.6, I've not looked at 0.6+ as yet, but should work there too, its trivial.
Patch works with mock-0.6.2 fine here. (:
-- Rex
--- mock.py.rpm 2006-09-26 05:08:41.000000000 -0700 +++ mock.py 2006-09-26 05:09:08.000000000 -0700 @@ -643,6 +643,7 @@ for item in [self.basedir, self.rootdir, self.statedir, self.resultdir, os.path.join(self.rootdir, 'var/lib/rpm'), os.path.join(self.rootdir, 'var/log'), + os.path.join(self.rootdir, 'var/lock/rpm'), os.path.join(self.rootdir, 'dev'), os.path.join(self.rootdir, 'etc/rpm'), os.path.join(self.rootdir, 'tmp'),
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Jeff Sheltren wrote:
Thanks, Rex. I'm also happy with it on my mock 0.4 installs, and haven't seen any ill effects :)
So, any objections to me committing the attached patch? It is simply Karanbir's patch adapted for current CVS.
Thanks, Jeff
Yeah, that patch (or my version of it) is already in CVS.
Clark
Cool, thanks - it wasn't there when I did a checkout this morning ;)
-Jeff
Clark Williams wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Jeff Sheltren wrote:
Thanks, Rex. I'm also happy with it on my mock 0.4 installs, and haven't seen any ill effects :)
So, any objections to me committing the attached patch? It is simply Karanbir's patch adapted for current CVS.
Thanks, Jeff
Yeah, that patch (or my version of it) is already in CVS.
Clark
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Jeff Sheltren wrote:
Cool, thanks - it wasn't there when I did a checkout this morning ;)
Well, it *should* have been...
/me roots around in his mock HEAD sandbox...
GAH! there it sits, not committed.
In my defense, it's actually there in the FC6 branch. But I didn't commit it to the HEAD.
It's there now.
Mea Culpa...
Clark
Clark Williams wrote:
Well, it *should* have been...
/me roots around in his mock HEAD sandbox...
GAH! there it sits, not committed.
In my defense, it's actually there in the FC6 branch. But I didn't commit it to the HEAD.
It's there now.
Mea Culpa...
Clark
Hi Clark, thanks for your help!
-Jeff
buildsys@lists.fedoraproject.org