OK this
seems to work:
def testRunnerContext = context.getProperty("LoadTestRunner").context
def threadPool = testRunnerContext.getProperty("threadPool")
def currentThread = Thread.currentThread()
log.info "Thread " + currentThread.id + " starting, threadPool size = " + threadPool.size
if (threadPool.size() >= 4) {
threadPool.each { thread ->
thread.interrupt()
}
} else {
synchronized( this ) {
threadPool.add(currentThread)
}
log.info threadPool.size
try {
currentThread.sleep(10000);
}
catch (InterruptedException ie) {
currentThread.interrupt()
log.info "Thread " + currentThread.id + " interrupted."
}
}
log.info "Thread " + Thread.currentThread().id + " finished."But what are the risks of this approach? Also, it would be nice, if all the required threads did not reach the script, that all the remaining threads would wake up at the same time.
