hi all

I'm having a long list of errors in loggregator_tc stdout log file, saying :

{"timestamp":1433153574.809743166,"process_id":13854,"source":"loggregator
trafficcontroller","log_level":"error","message":"WebsocketListener.Start:
Error connecting to ws://
198.11.228.56:8081/apps/71725662-41b5-4b1f-a37f-da28b47ece75/stream:
websocket: close 1005
","data":null,"file":"/var/vcap/data/compile/loggregator_trafficcontroller/loggregator/src/trafficcontroller/listener/websocket_listen
er.go","line":73,"method":"trafficcontroller/listener.(*websocketListener).listenWithTimeout"}



found the source code responsible for the message:


https://github.com/gorilla/websocket/blob/master/conn.go
called from
https://github.com/cloudfoundry/loggregator/blob/develop/src/trafficcontroller/listener/websocket_listener.go



understood that 1005 is an error reporting to a connection being closed
missing the status information (as of RFC 6455, section 11.7)


// Close codes defined in RFC 6455, section 11.7.
const (
CloseNormalClosure = 1000
CloseGoingAway = 1001
CloseProtocolError = 1002
CloseUnsupportedData = 1003
CloseNoStatusReceived = 1005
CloseAbnormalClosure = 1006
CloseInvalidFramePayloadData = 1007
ClosePolicyViolation = 1008
CloseMessageTooBig = 1009
CloseMandatoryExtension = 1010
CloseInternalServerErr = 1011
CloseTLSHandshake = 1015
)


Each time a log is requested, this error is repeated for many of the
dopplers behind the traffic_controller (but rarely also for all of those),
and I'm still not able to understand if this should be a blocking error or
a sort of warning (I'm not receiving complaints from the users) due to
firehose connection method, but I need to know what is happening and how to
prevent any problem.

thx
GV