While provisioning a server with Ansible 1.9.1 from a OS X host, I get the following error at random points of the provisioning process:
No handlers could be found for logger "paramiko.transport"
fatal: [xxx.xxx.xxx.xxx] =>
{'msg': 'One or more items failed.', 'failed': True, 'changed': False, 'results':
[{'invocation': {'module_name': 'somemodule', 'module_args': ''},
'item': {'priv': 'ALL', 'moreitems'}, 'changed': False, 'user': ''},
{'msg': 'FAILED: Error reading SSH protocol banner[Errno
54] Connection reset by peer', 'failed': True}]}
FATAL: all hosts have already failed -- aborting
Why is the connection reset by peer at random points is beyond me. I have no configuration for paramiko logging what-so-ever.
If this bug on Ansible, on the linux server, on the OS X host, or somewhere else?
The problem ended up being Python 2.6 version of pycrypto.
The solution from dennen99 fixed it for me.
I'm assuming that upgrading to Python2.7 would also fix the problem (not tested by me).