4.2. Connection Viewer
This window gives a real-time view into all DataHub connections. You can open this window by clicking on the button in the Properties window.
The various columns identify the connection and show its status, with connection statistics, as follows:
Type The type of connection or thread. All connections of the same type will have the same type name.
Detail Some descriptive information about the individual connection. This can be different even for connections of the same type.
ID An internal ID that can distinguish otherwise identical connections.
Label A label for this connection, where available. Typically, outgoing connections are labelled as part of their configuration. Some connections may provide additional information in this label.
Status An indication of the status of the connection. This is dependent on the type of connection.
Sent The number of data point changes that have been delivered to the thread managing this connection. That is, the number of point changes that the data engine has emitted. This does not necessarily imply that these changes have been forwarded onward to the destination of this connection. For example, an MQTT client connection to an unavailable broker will still see its Sent count rise, but those point changes could be held internally in a queue until the broker is again available.
Received The number of data point changes that have been received from the thread managing this connection. This is typically the number of data point changes received from the underlying source, such as an OPC server.
Created The number of data points that are created in the DataHub data engine due to interaction with this connection. For example, a connection to an OPC server will create points in the DataHub instance to correspond to the OPC items in the server.
Registered The number of point registrations made by the thread managing this connection. This is an indication by the thread that it should receive point change notifications from the DataHub data engine. Those point changes will result in the Sent count rising.
Unregistered The number of times that the connection thread has unregistered a point in the data engine. This will change whenever the connection thread unregisters a point, even if it was not previously registered.
Dropped The DataHub program makes a trade-off between delivering all data changes vs. maintaining timely delivery. If a connection cannot keep up with the rate of data changes, the DataHub data engine will throttle data transmission by dropping data point changes that are known to be superseded by newer values for the same point. A data change will not be dropped if there is no newer value available for the same point. This ensures that a connection will receive the current value even if it misses intermediate values. Large numbers of dropped values indicate that the connection is unable to keep up with the current data rate.
Blocked, Unblocked Both the DataHub data engine and a connection thread monitor the data transmission rate, and determine whether the connection is able to maintain the data rate. If not, the connection is temporarily blocked, indicating to the data engine that it should begin dropping all intermediate values. When the connection is again able to deliver data, the connection will be unblocked. Frequent changes to the Blocked and Unblocked counts indicate that the connection is frequently becoming overwhelmed and then quickly recovering.
CPU (Native interface only) The amount of CPU used by the thread managing this connection. When the option is selected, the CPU percentage is the average percentage of the CPU resources that the thread has been using since the Connection Viewer window was opened, or the button was pressed.
(Native interface only) The operating system thread ID for this thread.
How many data point changes have been queued to the connection thread by the data engine but have not yet been consumed by the connection thread. If this number grows large and remains large, this typically indicates that the connection is being overwhelmed and does not have sufficient CPU available to handle the data changes. The data engine will eventually block future deliveries if this number becomes too large. Large queues can result in large process memory consumption.
The name of the user credential assigned to this connection. This typically is only available for incoming connections.
The number of kilobytes of data sent on this connection. This is typically only meaningful for TCP connections.
The number of kilobytes of data received from this connection. This is typically only meaningful for TCP connections.
You have the following options for viewing the data
Displays all connection statistics from the time that the client connection was first made.
Displays all connection statistics from the last time the button was clicked or from the time that the client started, whichever is later.
Shows the performance of the connection, in terms of data changes per second, from the last time the button was clicked or from the time that the client started, whichever is later.
Records the time and current statistics for each client, for use by the and options.