Methods for stopping the PATROL Agent

PATROL provides three methods to stop the agent, and each method uses a different PATROL interface:

• The command line
• A variety of PATROL utilities, particularly the pconfig +KILL option discussed in Using pconfig to configure the PATROL Agent
• The PATROL Developer Console (click the host icon, and select Developer > Kill Agent)

Stopping the PATROL process

When you stop the PATROL Agent, it performs the following processes:

• Writes the process cache
• Writes the history cache
• Waits for child processes to terminate, which may take more time to terminate
• Closes network connections
• Stops all other related processes

If you quit the PATROL Agent  from an OS shutdown script, you must introduce sleep logic to ensure that the Agent completely stops prior to the OS shutdown because the Agent waits for child processes to end.

Improper shutdown

Ending the agent improperly can cause problems including corruption of history files and event log files. Following are common methods of improperly stopping the PATROL Agent:

• The computer crashes.
• The computer is shutdown properly, but the PATROL Agent is still running.
• The PATROL Agent process is interrupted using the 'kill -9' command which forces the agent to stop without giving it a chance to close its files.

For information about repairing corrupt history databases, see Fixing a corrupted history database.

PATROL configuration utility (pconfig)

The advantages of stopping the agent with the PATROL configuration utility include:

• It does not require system administrator or root access.
• It provides a standard method that works across all platforms.

Syntax

The basic syntax of the stop command for the pconfig utility is as follows:

pconfig +KILL [options]

Options

The following table lists the options available for +KILL.

Where to go from here

Starting and stopping the PATROL Agent