Meta-issue - unexpected timeouts encountered during transfers

[fviard]

There is a long list of issues reported regarding tasks that are failing because of timeouts, or multiple retries failing:

Maybe duplicated tickets:
#226
#737
#819
#877
#907
#912
#913

Errors are like:
File "/usr/lib64/python2.6/ssl.py", line 136, in read
return self._sslobj.read(len)
error: [Errno 104] Connection reset by peer

File "/usr/lib/python2.7/httplib.py", line 836, in connect
self.timeout, self.source_address)
File "/usr/lib/python2.7/socket.py", line 575, in create_connection
raise err
error: [Errno 110] Operation timed out

[fviard]

When amazon timeout, for example because the data uploaded is shorter than expected:
ERROR: S3 error: 400 (RequestTimeout): Your socket connection to the server was not read from or written to within the timeout period. Idle connections will be closed.