Bug #44627
Updated by Ernesto Puerta about 4 years ago
In master (and Octopus), master, the following exception appears when booting ceph-mgr: <pre> 2020-03-16T17:02:12.313+0000 7fdc97340700 -1 mgr load Traceback (most recent call last): File "/ceph/src/pybind/mgr/crash/module.py", line 37, in __init__ super(Module, self).__init__(*args, **kwargs) File "/ceph/src/pybind/mgr/mgr_module.py", line 723, in __init__ log_level = self.get_module_option("log_level") File "/ceph/src/pybind/mgr/mgr_module.py", line 1216, in get_module_option self._validate_module_option(key) File "/ceph/src/pybind/mgr/mgr_module.py", line 1198, in _validate_module_option format(key, self.__class__.__name__)) RuntimeError: Config option 'log_level' is not in Module.MODULE_OPTIONS </pre> This is caused by "a recent refactoring":https://github.com/ceph/ceph/pull/33897/files#diff-af4be246b95af4d67b3fc62e30a49178L602-L603, which moved the logging options definition out of the initializer function of the MgrModule class, so now they happen after they are first [tried to be] used.