Univention Bugzilla – Bug 33963
Tracebacks in UVMM
Last modified: 2015-02-04 15:56:15 CET
There are tracebacks in UVMM that may be explained a bit more in detail what went wrong. Tracebacks give very weak user experience. File "/usr/lib/pymodules/python2.6/notifier/threads.py", line 82, in _run tmp = self._function() File "/usr/lib/pymodules/python2.6/notifier/__init__.py", line 104, in __call__ return self._function( *tmp, **self._kwargs ) File "/usr/lib/pymodules/python2.6/univention/management/console/modules/uvmm/uvmmd.py", line 84, in request raise UVMM_Error(str(ex)) UVMM_Error: Could not open socket "/var/run/uvmm.socket": 111 ====================================== Traceback: File "/usr/lib/pymodules/python2.6/notifier/threads.py", line 82, in _run tmp = self._function() File "/usr/lib/pymodules/python2.6/notifier/__init__.py", line 104, in __call__ return self._function( *tmp, **self._kwargs ) File "/usr/lib/pymodules/python2.6/univention/management/console/modules/uvmm/uvmmd.py", line 84, in request raise UVMM_Error(str(ex)) UVMM_Error: Error while waiting for answer: 104 ================= See also Bug #33748
(In reply to Dirk Wiesenthal from comment #0) > There are tracebacks in UVMM that may be explained a bit more in detail what > went wrong. ... > File "/usr/lib/pymodules/python2.6/univention/management/console/modules/uvmm/uvmmd.py", line 84, in request > raise UVMM_Error(str(ex)) > UVMM_Error: Could not open socket "/var/run/uvmm.socket": 111 The UMC-Middle-End-Python-Module is unable to talk to the UVMMd-Python-Backend. This can be caused by UVMMd no longer be running or a performance problem, as I just encountered with Bug #33458. > File "/usr/lib/pymodules/python2.6/univention/management/console/modules/uvmm/uvmmd.py", line 84, in request > raise UVMM_Error(str(ex)) > UVMM_Error: Error while waiting for answer: 104 Same category: Something went wrong with IPC. "Maybe UVMMd needs to be restarted?" Most probably caused by "Bug #33741" not doing its job right.
First one (UVMM_Error: Could not open socket "/var/run/uvmm.socket": 111) happened again. Reported via traceback feedback.
(In reply to Dirk Wiesenthal from comment #2) > First one (UVMM_Error: Could not open socket "/var/run/uvmm.socket": 111) > happened again. Reported via traceback feedback. But 3.2-0 errata16 (Borgfeld)... maybe actually fixed with Bug #33458?
*** Bug 24776 has been marked as a duplicate of this bug. ***
Reported again (the "Could not open socket" traceback). Means that it is not fixed yet.
reported again: UVMM_Error: Could not open socket "/var/run/uvmm.socket": 111 Remark: when i control my slave UCS in virtualization it get not responding
Also "UVMM_Error: Error while waiting for answer: 104" was reported again.
I was able to reproduce it in the following way: - install UVMMd - open the module and create a EC2 connection - let the module open and restart the uvmmd process via the system services module I got this traceback: File "/usr/lib/pymodules/python2.7/notifier/threads.py", line 82, in _run tmp = self._function() File "/usr/lib/pymodules/python2.7/notifier/__init__.py", line 104, in __call__ return self._function( *tmp, **self._kwargs ) File "/usr/lib/pymodules/python2.7/univention/management/console/modules/uvmm/uvmmd.py", line 84, in request raise UVMM_Error(str(ex)) UVMM_Error: Could not open socket "/var/run/uvmm.socket": 2
I had the first traceback. Steps were: * browser had an old UMC session (>8 hours) displaying the menu * click on the UVMM module opened the module and the authentication window * after authentication the traceback appeared UCS 4.0 RC
>>> os.strerror(2) 'No such file or directory' >>> os.strerror(104) 'Connection reset by peer' >>> os.strerror(111) 'Connection refused'
Reported again.
Reported with UCS 4.0: UVMM_Error: Could not open socket "/var/run/uvmm.socket": 2
errno 2 reported again with 4.0-0 errata10 (Walle).
(In reply to Florian Best from comment #13) > errno 2 reported again with 4.0-0 errata10 (Walle). twice
Reported: 4.0-0 errata0 (Walle)
The ideal solution would be to display the servers / VM's as "not available" when this error occurs.
(In reply to Florian Best from comment #16) > The ideal solution would be to display the servers / VM's as "not available" > when this error occurs. I do not think that this is ideal. If a server is down it is currently shown as 'not available'. To fix the problem i would debug the problem on that server. If all servers are shown as offline i would think of it as a network problem. We should show the error - the module could not connect to the local uvmm daemon - explicitly. Suggestion: Show an otherwise hidden widget above the tree and grid, do not show anything in the tree+grid.
Reported again, 4.0-0 errata21 (Walle)
Reported again, 4.0-0 errata41 (Walle)
Errno 2, 111 and 104 are now handled by displaying a human readable error message. Fix: svn r57705 YAML: 2015-02-02-univention-virtual-machine-manager-daemon.yaml Package: univention-virtual-machine-manager-daemon Version: 4.0.23-1.577.201502021347
This is a great user experience improvement. As discussed, one thing though: If the UMC module is opened if uvmmd is not running, the tree stays in an endless loading loop, even if uvmmd is started. A reload of the tree should be done if the connection to uvmmd comes back.
I added some workarounds / tree reloading for that case. It's currently not possible to do this automatically in every case. Package builds.
Reopen: As discussed, there are still some issues with the tree not reloading properly when clicking on the search button
Well, okay. I now reload the tree everytime after a successful search if it is empty.
OK: Error message OK: Tree reload OK: r57705, r57733, r57736, r57738 Fixed: yaml (r57740) Verified
<http://errata.univention.de/ucs/4.0/74.html>