Hide Forgot
Created attachment 523183 [details] vdsm log Description of problem: Thread-825::ERROR::2011-09-14 17:52:02,018::guestIF::268::vm.Vm::(run) vmId=`ff1ef43c-bf3d-4231-8545-26731306a048`::Unexpected exception: Traceback (most recent call last): File "/usr/share/vdsm/guestIF.py", line 265, in run self.log.exception("Run exception: %s", line) UnboundLocalError: local variable 'line' referenced before assignment while not self._stopped: try: line = self._readLine() # line is always None after stop() is called and the # socket is closed. if line: (message, args) = self._parseLine(line) self._agentTimestamp = time.time() self._handleMessage(message, args) except: self.log.exception("Run exception: %s", line) should set line to None before the read Version-Release number of selected component (if applicable): vdsm-4.9-98.el6.x86_64 How reproducible: happened in migration destenation Actual results: Expected results: Additional info:
BTW David, this bug was introduced in vdsm-4.9-99-7-ga5f8785 (vdsm-4.9-100). Any word on how to reproduce it, and the functional effect? Thanks for the suggested fix, I've posted it under your name at http://gerrit.usersys.redhat.com/936
(In reply to comment #2) > BTW David, this bug was introduced in vdsm-4.9-99-7-ga5f8785 (vdsm-4.9-100). > > Any word on how to reproduce it, and the functional effect? > > Thanks for the suggested fix, I've posted it under your name at > > http://gerrit.usersys.redhat.com/936 i don't know if there is any functional effects, and i don't know exactly how to reproduce it, i just saw it in the log after some concurrent migrations - since the fix is simple i didn't find it significant to investigate the cause. i can try to reproduce it anytime if needed.
If not fixed, vdsm may stop waiting for guest messages after migration (which is bad). This is a failure of the first implementation of the fix for bug 736422, and should be verified as a part of it. No need to chase more flags. *** This bug has been marked as a duplicate of bug 736422 ***