hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "stack (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-1386) NPE in housekeeping
Date Thu, 07 May 2009 16:49:32 GMT

    [ https://issues.apache.org/jira/browse/HBASE-1386?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12706989#action_12706989
] 

stack commented on HBASE-1386:
------------------------------

{code}
Index: src/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java
===================================================================
--- src/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java    (revision 772492)
+++ src/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java    (working copy)
@@ -1124,6 +1124,7 @@
     }
     return true;
   }
+
   /*
    * Run some housekeeping tasks before we go into 'hibernation' sleeping at
    * the end of the main HRegionServer run loop.
@@ -1132,12 +1133,16 @@
     // If the todo list has > 0 messages, iterate looking for open region
     // messages. Send the master a message that we're working on its
     // processing so it doesn't assign the region elsewhere.
-    if (this.toDo.size() <= 0) {
+    if (this.toDo.isEmpty()) {
       return;
     }
     // This iterator is 'safe'.  We are guaranteed a view on state of the
     // queue at time iterator was taken out.  Apparently goes from oldest.
     for (ToDoEntry e: this.toDo) {
+      HMsg msg = e.msg;
+      if (msg == null) {
+        LOG.warn("Message is empty: " + e);
+      }
       if (e.msg.isType(HMsg.Type.MSG_REGION_OPEN)) {
         addProcessingMessage(e.msg.getRegionInfo());
       }
{code}

> NPE in housekeeping
> -------------------
>
>                 Key: HBASE-1386
>                 URL: https://issues.apache.org/jira/browse/HBASE-1386
>             Project: Hadoop HBase
>          Issue Type: Bug
>            Reporter: stack
>            Assignee: stack
>             Fix For: 0.20.0
>
>
> Two regionservers hosting thousands of regions.  One went down. Other was trying to open
about 1000 regions and got this:
> {code}
> 2009-05-07 10:24:59,995 [regionserver/0:0:0:0:0:0:0:0:60021.worker] INFO org.apache.hadoop.hbase.regionserver.HRegionServer:
Worker: MSG_REGION_OPEN: REGION => {NAME => 'TestTable,1094797576,1241684896885', STARTKEY
=> '1094797576', ENDKEY => '1096819542', ENCOD
> ED => 1333893125, TABLE => {{NAME => 'TestTable', FAMILIES => [{NAME =>
'info', COMPRESSION => 'NONE', VERSIONS => '3', LENGTH => '2147483647', TTL =>
'-1', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCKCACHE => 'false'}]}}
> 2009-05-07 10:24:59,996 [regionserver/0:0:0:0:0:0:0:0:60021] FATAL org.apache.hadoop.hbase.regionserver.HRegionServer:
Unhandled exception. Aborting...
> java.lang.NullPointerException
>     at org.apache.hadoop.hbase.regionserver.HRegionServer.housekeeping(HRegionServer.java:1141)
>     at org.apache.hadoop.hbase.regionserver.HRegionServer.run(HRegionServer.java:518)
>     at java.lang.Thread.run(Thread.java:619)
> {code}

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message