6.11.7. Adjusting the Bridge Mode Forced Client Disconnect Timeout

This feature controls how long the Connector in Bridge mode waits before forcibly disconnecting the server side of the session after a client session ends.

Default: 500ms

When a client application opens a socket and connects to the connector, a second socket/connection to the server is created. The Connector in bridge mode then simply transfers data between these two sockets. 

When a client application brutally closes a connection without following the proper disconnection protocol, the server will not know about that disconnect until the connector closes the Connector<>server socket. If the connector closes the Connector<>server connection too soon after a client disconnect, there is a chance that the proper disconnection messages will be missed, if sent late. If the connector does not close this connector<>server connection, it would stay open indefinitely, using memory and resources that would otherwise be reclaimed. The default of 500ms is very conservative and will fit most environments where client applications disconnect properly. When the volume of connections opened and never closed exceeds a certain level, the timeout must be tuned (lowered) to close idle connections faster, or the available resources will get used up.

Many times this situation is caused by health checks, especially from monitoring scripts and load balancers checking port liveness. Many of these check do not gracefully close the connection, triggering the need for tuning the Connector.

If connections still exist after the timeout interval, they are forced closed, and a warning will be printed in the connector logs (C>S ended. S>C streaming did not finish within bridgeServerToClientForcedCloseTimeout=500 (ms). Will be closed anyway !).

Important

This setting ONLY applies to Bridge mode.

This timeout is adjusted via the tpm property --property=bridgeServerToClientForcedCloseTimeout in milliseconds.

For example, to change the delay to 50 milliseconds:

Click the link below to switch examples between Staging and INI methods...

Show Staging

Show INI

shell> tpm query staging
tungsten@db1:/opt/continuent/software/tungsten-clustering-6.0.5-41

shell> echo The staging USER is `tpm query staging| cut -d: -f1 | cut -d@ -f1`
The staging USER is tungsten

shell> echo The staging HOST is `tpm query staging| cut -d: -f1 | cut -d@ -f2`
The staging HOST is db1

shell> echo The staging DIRECTORY is `tpm query staging| cut -d: -f2`
The staging DIRECTORY is /opt/continuent/software/tungsten-clustering-6.0.5-41

shell> ssh {STAGING_USER}@{STAGING_HOST}
shell> cd {STAGING_DIRECTORY}
shell> ./tools/tpm configure alpha \
    --property=bridgeServerToClientForcedCloseTimeout=50

Run the tpm command to update the software with the Staging-based configuration:

shell> ./tools/tpm update

For information about making updates when using a Staging-method deployment, please see Section 9.3.7, “Configuration Changes from a Staging Directory”.

shell> vi /etc/tungsten/tungsten.ini
[alpha]
...
property=bridgeServerToClientForcedCloseTimeout=50

Run the tpm command to update the software with the INI-based configuration:

shell> tpm query staging
tungsten@db1:/opt/continuent/software/tungsten-clustering-6.0.5-41

shell> echo The staging DIRECTORY is `tpm query staging| cut -d: -f2`
The staging DIRECTORY is /opt/continuent/software/tungsten-clustering-6.0.5-41

shell> cd {STAGING_DIRECTORY}

shell> ./tools/tpm update

For information about making updates when using an INI file, please see Section 9.4.4, “Configuration Changes with an INI file”.

Warning

If you decrease this value, you run the risk of disconnecting valid but slow sessions.

Important

Updating these values require a connector restart (via tpm update) for the changes to be recognized.