Red Hat Bugzilla – Bug 601025
beaker's log couldn't prompt enough information to determine where the problem is
Last modified: 2010-12-20 10:30:12 EST
Description of problem:
I could get obvious problem prompt when run rhts testcase by rhts log, but beaker's log couldn't give me like that.
pls refer to below links:
I got the error info:
and could get clear info by check the log, so I search and find it's a bug opened:
libvirtError: internal error Process exited while reading console log output: char device redirected to /dev/pts/2
qemu: could not open disk image /var/lib/libvirt/images/guest-81-163.rhts.eng.bos.redhat.com: Invalid argument
but in beaker, just job abort and no clear info shown where the problem like rhts behaviour:
clear error info prompt.
Seems like a regression. Missing debugging information for test can have a significant impact on QE productivity.
I suggest to make this a blocker.
I think this is related to having NetworkManager installed.
Right now the kickstarts are different between rhts and beaker. This will be fixed in the next upgrade.
The point is though, that the logging is working, we are not getting to the actual install of the virt guest because of something going wrong with bringing up the bridge interface.
If someone could log into the system when this is happening it would be useful.
I'll try and debug if I get a chance, but I'm pretty busy. :-)
I have seen this happening on snapshot6...
See Bug 598631 - shutdown, reboot, halt and C-A-D don't work
= Workaround =
1. Bug 598631: connect to the machine (console/ssh) and run
2. Bug 589904: after reboot run
pgrep beah-srv || rm -f /var/beah/*.socket && service beah-srv start
This is fixed upstream and will be delivered soon.