[HBASE-3062] ZooKeeper KeeperException$ConnectionLossException is a "recoverable" exception; we should retry a while on server startup at least.

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.90.0
• Hadoop Flags: Reviewed
• Assignee: Michael Stack
• Reporter: Michael Stack
• Created: 01/Oct/10 04:59
• Updated: 20/Nov/15 12:43
• Resolved: 09/Oct/10 17:55

Description

On startup of daemons we'll sometimes fail (new master) with something like the following:

2010-09-30 23:38:36,979 INFO org.apache.hadoop.hbase.zookeeper.ZKUtil: regionserver60020 opening connection to ZooKeeper with quorum (sv2borg182:20001,sv2borg181:20001,sv2borg180:20001)
2010-09-30 23:38:36,979 INFO org.apache.zookeeper.ZooKeeper: Initiating client connection, connectString=sv2borg182:20001,sv2borg181:20001,sv2borg180:20001 sessionTimeout=60000 watcher=regionserver60020
2010-09-30 23:38:36,980 INFO org.apache.zookeeper.ClientCnxn: Opening socket connection to server sv2borg182/10.20.20.182:20001
2010-09-30 23:38:36,980 INFO org.apache.zookeeper.ClientCnxn: Socket connection established to sv2borg182/10.20.20.182:20001, initiating session
2010-09-30 23:38:36,981 INFO org.apache.zookeeper.ClientCnxn: Unable to read additional data from server sessionid 0x0, likely server has closed socket, closing socket connection and attempting reconnect
2010-09-30 23:38:37,083 ERROR org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher: regionserver60020 Unexpected KeeperException creating base node
org.apache.zookeeper.KeeperException$ConnectionLossException: KeeperErrorCode = ConnectionLoss for /hbase
    at org.apache.zookeeper.KeeperException.create(KeeperException.java:90)
    at org.apache.zookeeper.KeeperException.create(KeeperException.java:42)
    at org.apache.zookeeper.ZooKeeper.create(ZooKeeper.java:637)
    at org.apache.hadoop.hbase.zookeeper.ZKUtil.createAndFailSilent(ZKUtil.java:807)
    at org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher.<init>(ZooKeeperWatcher.java:107)
    at org.apache.hadoop.hbase.regionserver.HRegionServer.initializeZooKeeper(HRegionServer.java:438)
    at org.apache.hadoop.hbase.regionserver.HRegionServer.initialize(HRegionServer.java:420)
    at org.apache.hadoop.hbase.regionserver.HRegionServer.<init>(HRegionServer.java:305)
    at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
    at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:39)
    at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:27)
    at java.lang.reflect.Constructor.newInstance(Constructor.java:513)
    at org.apache.hadoop.hbase.regionserver.HRegionServer.constructRegionServer(HRegionServer.java:2436)
    at org.apache.hadoop.hbase.regionserver.HRegionServerCommandLine.start(HRegionServerCommandLine.java:60)
    at org.apache.hadoop.hbase.regionserver.HRegionServerCommandLine.run(HRegionServerCommandLine.java:75)
    at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:65)
    at org.apache.hadoop.hbase.util.ServerCommandLine.doMain(ServerCommandLine.java:76)
    at org.apache.hadoop.hbase.regionserver.HRegionServer.main(HRegionServer.java:2460)

I'll see it over on master too the odd time.

Currently we fail out (though in above case, because of order in which we do startup the regionserver actually hung up because RPC was started before zk)

We should retry this error some at least on startup because its 'recoverable' (See http://wiki.apache.org/hadoop/ZooKeeper/ErrorHandling). In fact, we should probably retry always until we get a session expired.