Update scripts so that container exits promptly after server shuts down

Author: rjeberhard

operator/src/main/resources/scripts/startServer.sh

@@ -25,7 +25,7 @@ trace "Starting WebLogic Server '${SERVER_NAME}'."
 function exitOrLoop {
   if [ -f /weblogic-operator/debug/livenessProbeSuccessOverride ]
   then
-    while true ; do sleep 60 ; done
+    waitForShutdownMarker
   else
     exit 1
   fi
@@ -100,23 +100,21 @@ function waitUntilShutdown() {
   #
   if [ "${SERVER_OUT_IN_POD_LOG}" == 'true' ] ; then
     trace "Showing the server out file from ${SERVER_OUT_FILE}"
-    tail -F -n +0 ${SERVER_OUT_FILE} || exitOrLoop
-  else
-    trace "Wait indefinitely so that the Kubernetes pod does not exit and try to restart"
-    while true; do sleep 60; done
+    ${SCRIPTPATH}/tailLog.sh ${SERVER_OUT_FILE} &
   fi
+  waitForShutdownMarker
 }
 
-function mockWaitUntilShutdown() {
+function waitForShutdownMarker() {
   #
   # Wait forever.   Kubernetes will monitor this pod via liveness and readyness probes.
   #
   trace "Wait indefinitely so that the Kubernetes pod does not exit and try to restart"
   while true; do
-    if [ -e /u01/doShutdown ] ; then
+    if [ -e /weblogic-operator/doShutdown ] ; then
       exit 0
     fi
-    sleep 5
+    sleep 3
   done
 }
 
@@ -219,7 +217,7 @@ copySitCfg /weblogic-operator/introspector ${DOMAIN_HOME}/optconfig/diagnostics
 
 if [ "${MOCK_WLS}" == 'true' ]; then
   mockWLS
-  mockWaitUntilShutdown
+  waitForShutdownMarker
 else
   startWLS
   waitUntilShutdown

operator/src/main/resources/scripts/stopServer.sh

@@ -13,12 +13,12 @@ SCRIPTPATH="$( cd "$(dirname "$0")" > /dev/null 2>&1 ; pwd -P )"
 source ${SCRIPTPATH}/traceUtils.sh
 [ $? -ne 0 ] && echo "Error: missing file ${SCRIPTPATH}/traceUtils.sh" && exit 1
 
-trace "Stop server ${SERVER_NAME}" &>> /u01/oracle/stopserver.out
+trace "Stop server ${SERVER_NAME}" &>> /weblogic-operator/stopserver.out
 
 checkEnv SERVER_NAME || exit 1
 
 if [ "${MOCK_WLS}" == 'true' ]; then
-  touch /u01/doShutdown
+  touch /weblogic-operator/doShutdown
   exit 0
 fi
 
@@ -28,27 +28,27 @@ function check_for_shutdown() {
   state=`${SCRIPTPATH}/readState.sh`
   exit_status=$?
   if [ $exit_status -ne 0 ]; then
-    trace "Node manager not running or server instance not found; assuming shutdown" &>> /u01/oracle/stopserver.out
+    trace "Node manager not running or server instance not found; assuming shutdown" &>> /weblogic-operator/stopserver.out
     return 0
   fi
 
   if [ "$state" = "SHUTDOWN" ]; then
-    trace "Server is shutdown" &>> /u01/oracle/stopserver.out
+    trace "Server is shutdown" &>> /weblogic-operator/stopserver.out
     return 0
   fi
 
   if [[ "$state" =~ ^FAILED ]]; then
-    trace "Server in failed state" &>> /u01/oracle/stopserver.out
+    trace "Server in failed state" &>> /weblogic-operator/stopserver.out
     return 0
   fi
 
-  trace "Server is currently in state $state" &>> /u01/oracle/stopserver.out
+  trace "Server is currently in state $state" &>> /weblogic-operator/stopserver.out
   return 1
 }
 
 # Check if the server is already shutdown
 check_for_shutdown
-[ $? -eq 0 ] && trace "Server already shutdown or failed" &>> /u01/oracle/stopserver.out && exit 0
+[ $? -eq 0 ] && trace "Server already shutdown or failed" &>> /weblogic-operator/stopserver.out && exit 0
 
 # Otherwise, connect to the node manager and stop the server instance
 [ ! -f "${SCRIPTPATH}/wlst.sh" ] && trace "Error: missing file '${SCRIPTPATH}/wlst.sh'." && exit 1
@@ -60,28 +60,23 @@ export SHUTDOWN_TIMEOUT_ARG=${SHUTDOWN_TIMEOUT:-30}
 export SHUTDOWN_IGNORE_SESSIONS_ARG=${SHUTDOWN_IGNORE_SESSIONS:-false}
 export SHUTDOWN_TYPE_ARG=${SHUTDOWN_TYPE:-Graceful}
 
-trace "Before stop-server.py [${SERVER_NAME}] ${SCRIPTDIR}" &>> /u01/oracle/stopserver.out
-${SCRIPTPATH}/wlst.sh /weblogic-operator/scripts/stop-server.py &>> /u01/oracle/stopserver.out
-trace "After stop-server.py" &>> /u01/oracle/stopserver.out
+trace "Before stop-server.py [${SERVER_NAME}] ${SCRIPTDIR}" &>> /weblogic-operator/stopserver.out
+${SCRIPTPATH}/wlst.sh /weblogic-operator/scripts/stop-server.py &>> /weblogic-operator/stopserver.out
+trace "After stop-server.py" &>> /weblogic-operator/stopserver.out
 
 # at this point node manager should have terminated the server
 # but let's try looking for the server process and
 # kill the server if the process still exists,
 # just in case we failed to stop it via wlst
 pid=$(jps -v | grep " -Dweblogic.Name=${SERVER_NAME} " | awk '{print $1}')
 if [ ! -z $pid ]; then
-  echo "Killing the server process $pid" &>> /u01/oracle/stopserver.out
+  echo "Killing the server process $pid" &>> /weblogic-operator/stopserver.out
   kill -15 $pid
 fi
 
-# stop node manager process
-#
-trace "Stopping NodeManager" &>> /u01/oracle/stopserver.out
-pid=$(jps | grep "weblogic.NodeManager" | awk '{print $1}')
-trace "PID=[${pid}]" &>> /u01/oracle/stopserver.out
-if [ ! -z $pid ]; then
-  echo "Killing NodeManager process $pid" &>> /u01/oracle/stopserver.out
-  kill -9 $pid
+touch /weblogic-operator/doShutdown
+if [ -f /weblogic-operator/pid ]; then
+  kill -2 $(</weblogic-operator/pid)
 fi
 
-trace "Exit script"  &>> /u01/oracle/stopserver.out
+trace "Exit script"  &>> /weblogic-operator/stopserver.out

operator/src/main/resources/scripts/tailLog.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+
+# Copyright 2019, Oracle Corporation and/or its affiliates.  All rights reserved.
+# Licensed under the Universal Permissive License v 1.0 as shown at
+# http://oss.oracle.com/licenses/upl.
+
+#
+# This script is used to tail the server log and is separate from
+# startServer.sh so that it is easier to quickly kill the process
+# running this script.
+#
+
+echo $$ > /weblogic-operator/pid
+tail -F -n +0 $1

operator/src/test/java/oracle/kubernetes/operator/helpers/ConfigMapHelperTest.java

@@ -68,7 +68,8 @@ public class ConfigMapHelperTest {
     "startNodeManager.sh",
     "traceUtils.py",
     "traceUtils.sh",
-    "wlst.sh"
+    "wlst.sh",
+    "tailLog.sh"
   };
 
   private static final String introspectResult =