Hi,
I'm trying to get some insight into bug 1315541, where the F24 KDE live image ext4 rootfs is corrupt. What I'd like is the complete systemd journal for the VM. Is that somewhere? It might reveal if maybe something unique to KDE is hanging and then systemd hits a wait timeout and shuts down anyway (probably not, that'd be bad).
And then also I want to find out what lmc is using as squashfs arguments to pass to mksquashfs. When I use mksquashfs on an image with identical contents as the live image rootfs.img, I end up with a file that's 15% bigger than the compose process's squashfs.img.
I used to find some logs buried in the koji task for these but either since branch, or pungi4, the ant trail has been smeared so I can't find my way back to the picnic basket at the moment.
Thanks,
On Thu, 2016-03-10 at 14:22 -0700, Chris Murphy wrote:
Hi,
I'm trying to get some insight into bug 1315541, where the F24 KDE live image ext4 rootfs is corrupt. What I'd like is the complete systemd journal for the VM. Is that somewhere? It might reveal if maybe something unique to KDE is hanging and then systemd hits a wait timeout and shuts down anyway (probably not, that'd be bad).
And then also I want to find out what lmc is using as squashfs arguments to pass to mksquashfs. When I use mksquashfs on an image with identical contents as the live image rootfs.img, I end up with a file that's 15% bigger than the compose process's squashfs.img.
I used to find some logs buried in the koji task for these but either since branch, or pungi4, the ant trail has been smeared so I can't find my way back to the picnic basket at the moment.
They're right in the compose tree.
https://kojipkgs.fedoraproject.org/compose/branched/Fedora-24-20160310.n.0/l...
It's a bit of a guessing game which arch directory to find the log you want in...live compose logs are in i386-x86_64. -- Adam WilliamsonFedora QA Community MonkeyIRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . nethttp://www.happyassassin.net
On Thursday, March 10, 2016 02:22:30 PM Chris Murphy wrote:
Hi,
I'm trying to get some insight into bug 1315541, where the F24 KDE live image ext4 rootfs is corrupt. What I'd like is the complete systemd journal for the VM. Is that somewhere? It might reveal if maybe something unique to KDE is hanging and then systemd hits a wait timeout and shuts down anyway (probably not, that'd be bad).
And then also I want to find out what lmc is using as squashfs arguments to pass to mksquashfs. When I use mksquashfs on an image with identical contents as the live image rootfs.img, I end up with a file that's 15% bigger than the compose process's squashfs.img.
I used to find some logs buried in the koji task for these but either since branch, or pungi4, the ant trail has been smeared so I can't find my way back to the picnic basket at the moment.
Thanks,
The logs from the livemedia-tasks are in koji. http://koji.fedoraproject.org/koji/tasks?owner=releng&state=all&view... will get you the tasks. Note that we use livemedia-creator in non virt mode, there is no vm systemd journal to fetch. https://github.com/rhinstaller/lorax/pull/82 may help you understand what lorax is doing better.
For the things made by pungi the logs are in the compose tree. And in the case of the lives https://kojipkgs.fedoraproject.org/compose/rawhide/Fedora-Rawhide-20160310.n... the log points to the koji task where to get the logs so the first place you should look is in the compose directory. https://kojipkgs.fedoraproject.org/compose/
Dennis