From 508cb1bd987530790290dd6ec33bc6007bcc7f27 Mon Sep 17 00:00:00 2001
From: Michael Stack <stack@apache.org>
Date: Thu, 18 Jul 2013 17:01:44 +0000
Subject: [PATCH] HBASE-8979 JVMClusterUtil will join forever on a stuck
 regionserver <--- zombie maker

git-svn-id: https://svn.apache.org/repos/asf/hbase/trunk@1504527 13f79535-47bb-0310-9956-ffa450edef68
---
 .../apache/hadoop/hbase/util/JVMClusterUtil.java   | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/util/JVMClusterUtil.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/util/JVMClusterUtil.java
index 149b057f81b..27f76adbf3e 100644
--- a/hbase-server/src/main/java/org/apache/hadoop/hbase/util/JVMClusterUtil.java
+++ b/hbase-server/src/main/java/org/apache/hadoop/hbase/util/JVMClusterUtil.java
@@ -251,16 +251,17 @@ public class JVMClusterUtil {
 
     }
     boolean wasInterrupted = false;
-    final long maxTime = System.currentTimeMillis() + 120 * 1000;
+    final long maxTime = System.currentTimeMillis() + 30 * 1000;
     if (regionservers != null) {
       // first try nicely.
       for (RegionServerThread t : regionservers) {
         t.getRegionServer().stop("Shutdown requested");
       }
       for (RegionServerThread t : regionservers) {
-        if (t.isAlive() && !wasInterrupted && System.currentTimeMillis() < maxTime) {
+        long now = System.currentTimeMillis();
+        if (t.isAlive() && !wasInterrupted && now < maxTime) {
           try {
-            t.join(maxTime);
+            t.join(maxTime - now);
           } catch (InterruptedException e) {
             LOG.info("Got InterruptedException on shutdown - " +
                 "not waiting anymore on region server ends", e);
@@ -270,17 +271,20 @@ public class JVMClusterUtil {
       }
 
       // Let's try to interrupt the remaining threads if any.
-      for (int i = 0; i < 10; ++i) {
+      for (int i = 0; i < 100; ++i) {
+        boolean atLeastOneLiveServer = false;
         for (RegionServerThread t : regionservers) {
           if (t.isAlive()) {
+            atLeastOneLiveServer = true;
             try {
               LOG.warn("RegionServerThreads remaining, give one more chance before interrupting");
-              t.join(10);
+              t.join(1000);
             } catch (InterruptedException e) {
               wasInterrupted = true;
             }
           }
         }
+        if (!atLeastOneLiveServer) break;
         for (RegionServerThread t : regionservers) {
           if (t.isAlive()) {
             LOG.warn("RegionServerThreads taking too long to stop, interrupting");