|
Issue
|
Cause
|
Solution
|
|---|---|---|
|
Client receiving authentication requests for each resource requested
|
Cause: Cookies are used, and there is a mix of HTTP and HTTPS in the configuration. ThingWorx returns a cookie with the Secure flag specified, meaning the cookie is only used for HTTPS requests. When the client makes non-secure HTTP requests to the Connection Server, the client does not send the cookie. The ThingWorx Platform receives the next request without a cookie and without credentials, so it sends back a 401 response. This happens for each request from the client
Solution: Always use HTTPS when sending requests through the Connection Server to the ThingWorx Platform.
|
|
|
Failed to Start - Port is in Use
|
If another service, or another instance of Connection Server is already running on the port that you defined in the protocol.port property, you will see a large stack trace upon startup with the message:
Caused by: java.net.BindException: Address already in use ...
Caused by: java.io.IOException: Keystore was tampered with, or password was incorrect...
Caused by: java.security.UnrecoverableKeyException: Password verification failed
|
Change the protocol.port property to a port that is not in use. Note that the default is port 8080.
Change the health-check.port property to a port that is not in use. Note that the default is port 9009.
|
|
Failed to Start – Keystore Password
|
Upon startup, when using TLS, the system terminates with a long stack trace containing the following error messages: ERROR c.t.c.ConnectionServer - Unable to start ConnectionServer
|
Check that the protocol.undertow.ssl.keystore.password property is set to the password that is used by the keystore defined by the protocol.undertow.ssl.keystore.file property.
|
|
Failed to Start - Bad ThingWorx Endpoint
|
Upon startup, the system hangs with either of these two exceptions:
ERROR c.t.f.i.AbstractManyPlatformWebSocketFabric - Failure while connecting to platform: endpoint=wss://<yourhost>:<yourport>/Thingworx/WS
ERROR c.t.s.i.t.netty.NettyBytesChannel - Error while establishing WebSocket connection on channel com.thingworx.sdk.impl.transport.netty.NettyBytesChannel$2@5dd7e3e6.java.nio.channels.UnresolvedAddressException: null
|
Make sure that the ThingWorx Platform is running as you expect at the host and port defined in the transport.websockets.platforms property.
Try hitting the http/https endpoint with a browser. If not available, start up the ThingWorx Platform.
|
|
Failed to Start – Wrong Protocol Used
|
Upon startup, the system, repeatedly throws the following error in a stack trace:
WARN c.t.s.i.t.netty.NettyChannelHandler - [ClientHandler: 2001739664] SSLHandshake Exception, websocket not created: io.netty.handler.ssl.NotSslRecordException: not an SSL/TLS record: ###################’s…
|
Check the property, transport.websockets.platforms, to see if it specifies ws://. If the ThingWorx platform is running TLS, change the property value to start with wss://.
|
|
Failed to Start – Wrong Protocol Used
|
During runtime, the system throws errors:
WARN c.t.s.i.t.netty.NettyChannelHandler - [ClientHandler: 2001739664] SSLHandshake Exception, websocket not created:io.netty.handler.ssl.NotSslRecordException: not an SSL/TLS record: ###################’s…
|
The property, transport.websockets.platforms, specifies a wss:// endpoint, but the ThingWorx Platform is not actually running TLS. Change the property value to start with ws://, or specify a different, secure endpoint.
|
|
After a restart, the Connection Server bind fails.
|
After a restart in a ThingWorx High Availability Clustering deployment, the ThingWorx Platform may take up to 2 or 3 minutes to bind the Connection Server. For v.9.0.0, the default bind retry setting is every 15 seconds until it succeeds.
The following messages appear in the Connection Server log if a bind fails:
2020-04-11 14:49:20,770 [cxthing-bind-0] INFO c.t.c.CXThingBindStateListener - Binding connection server name=cxserver1 to thing template=ConnectionServer 2020-04-11 14:49:20,803 [cxthing-bind-0] WARN c.t.c.CXThingBindStateListener - Failed to bind ConnectionServer thing: name=cxserver1, template=ConnectionServer, code=STATUS_INTERNAL_ERROR, message=Error trying to process message: Binding failed, see server log for more information V1BindMessage [header=V1MessageHeader [messageType=10, requestId=16, endpointId=-1717057734, sessionId=-1, flags=0], gatewayName=cxserver1, gatewayType=ConnectionServer, names=[cxserver1]]. Retrying bind in 15s When the bind finally succeeds, the following messages appear in the Connection Server log:
019-12-11 14:51:50,986 [cxthing-bind-0] INFO c.t.c.CXThingBindStateListener - Binding connection server name=cxserver1 to thing template=ConnectionServer 2019-12-11 14:51:51,048 [cxthing-bind-0] INFO c.t.c.CXThingBindStateListener - Successfully bound ConnectionServer thing: name=cxserver1, template=ConnectionServer |