Class ProcessKiller

  • All Implemented Interfaces:
    java.io.Serializable

    public abstract class ProcessKiller
    extends java.lang.Object
    implements java.io.Serializable
    Extension point that defines more elaborate way of killing processes, such as sudo or pfexec, for ProcessTree.

    Lifecycle

    Each implementation of ProcessKiller is instantiated once on the master. Whenever a process needs to be killed, those implementations are serialized and sent over to the appropriate slave, then the kill(ProcessTree.OSProcess) method is invoked to attempt to kill the process.

    One of the consequences of this design is that the implementation should be stateless and concurrent-safe. That is, the kill(ProcessTree.OSProcess) method can be invoked by multiple threads concurrently on the single instance.

    Another consequence of this design is that if your ProcessKiller requires configuration, it needs to be serializable, and configuration needs to be updated atomically, as another thread may be calling into kill(ProcessTree.OSProcess) just when you are updating your configuration.

    Since:
    1.362
    Author:
    jpederzolli, Kohsuke Kawaguchi
    See Also:
    Serialized Form
    • Constructor Summary

      Constructors 
      Constructor Description
      ProcessKiller()  
    • Method Summary

      All Methods Instance Methods Abstract Methods 
      Modifier and Type Method Description
      abstract boolean kill​(ProcessTree.OSProcess process)
      Attempts to kill the given process.
      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
    • Constructor Detail

      • ProcessKiller

        public ProcessKiller()
    • Method Detail

      • kill

        public abstract boolean kill​(ProcessTree.OSProcess process)
                              throws java.io.IOException,
                                     java.lang.InterruptedException
        Attempts to kill the given process.
        Parameters:
        process - process to be killed. Always a local process.
        Returns:
        true if the killing was successful, and Hudson won't try to use other ProcessKiller implementations to kill the process. false if the killing failed or is unattempted, and Hudson will continue to use the rest of the ProcessKiller implementations to try to kill the process.
        Throws:
        java.io.IOException - The caller will log this exception and otherwise treat as if the method returned false, and moves on to the next killer.
        java.lang.InterruptedException - if the callee performs a time consuming operation and if the thread is canceled, do not catch InterruptedException and just let it thrown from the method.