Alive mechanism of manager connections
The alive mechanism in WinCC OA serves for connection controlling between the single managers. Each manager sends alive messages at definable intervals.
The mechanism is used to detect connection failures between single managers.
TCP guarantees that a broken connection is detected. The time required to detect a broken connection is, however, very long. The alive telegrams guarantee that a broken connection is detected within the configured time.
Each manager in WinCC OA can send and receive alive messages. Alive messages are sent in a separate thread and are sent via a separate TCP connection so that a blocking manager is not considered as a broken connection. Additionally, each manager sends alive messages via the normal manager connection in order to detect a connection failure. Since this is done in the main thread, the alive message is not checked but the sender of the operating system is informed when a sending fails. By default all managers send alive messages to all connected managers that do not run on the same host.
In this configuration the event manager knows that at least one alive message is delivered every 10 seconds from the UI over the UDP port. Additionally, the UI sends over the TCP connection an alive message during the same time period in order to detect a possible connection failure.
Alive messages are configured by using the following entry in the config file:
aliveTimeout: The entry specifies at which intervals (in seconds) this manager sends alive messages. Actually about 10 alive messages are sent over this period. When building a connection, this value is transferred to another manager and this manager starts the monitoring of the alive timeout. Negative values mean that alive messages are only sent via a connection to the remote computer to a manager on the remote computer. The alive timeout is in this case the absolute value. Positive values mean that messages are sent via all connections also via connections to a local computer and local manager. You can also deactivate the alive messages for one manager of a connection and activate them for another one. The value applies to all connections of a manager.
alivePriortiyClass: This entry sets the priority under which the alive thread runs:
-
0 means the normal priority,
-
1 means high priority and better guarantee that the thread is executed regularly also when the computer is under heavy load.
Debug levels
The receiving of alive messages can be monitored via the debug outputs of the respective manager in the Log Viewer.
Debug level | Description |
---|---|
-rcv 1 | The receiving of alive messages over the TCP connection can be displayed via this option. There is a system message KEEP_ALIVE every few seconds. |
-dbg alive | Debug output for the sending and receiving of alive messages. |