Description of problem: Running virt-install from Chinese, Japanese or any other non english locale, ends up causing virt-install to fail prematurely. It bails out due to a bad date format ( as far as I can tell) in the virt-manager, which is passed through to python logging framework. Version-Release number of selected component (if applicable): F10, F11, Rawhide, RHEL5. How reproducible: Every time. Steps to Reproduce: 1. Become one with uid 0. 2. export LANG=zh_CN.UTF8 3.# virt-install Traceback (most recent call last): File "/usr/lib64/python2.6/logging/handlers.py", line 74, in emit if self.shouldRollover(record): File "/usr/lib64/python2.6/logging/handlers.py", line 145, in shouldRollover msg = "%s\n" % self.format(record) File "/usr/lib64/python2.6/logging/__init__.py", line 637, in format return fmt.format(record) File "/usr/lib64/python2.6/logging/__init__.py", line 428, in format s = self._fmt % record.__dict__ UnicodeDecodeError: 'ascii' codec can't decode byte 0xe4 in position 1: ordinal not in range(128) ERROR 虚拟机需要一个名称。 I assume the above message says like its english counterpart which is something along the lines of name not specified. Actual results: Application throws python traceback exception. Expected results: Application to not throw an error. Additional info: I will attach a patch which attempts to fix the problem, the only downside that I could notice is that the date stamps in the log file$HOME/.virtinstall/virt-install.log were no longer in the native locale. Error messages remained the same.
Created attachment 350583 [details] I'm not sure if this is the best way,but at least it starts. I found a way to work around this bug by modifying the loggers __init__.py although I'd imagine that it would require considerable QE in comparison to this work around.
Thanks for the patch. There are actually other problems in this area that need a more complete fix, which are being tracked in bug 508784. *** This bug has been marked as a duplicate of bug 508784 ***
Yeah, I figured that was the case. I have a fix for the logger too, although the solution was a little more complex. Will transfer IT to this bug, Thanks Cole.