accumulo-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Eric Newton <eric.new...@gmail.com>
Subject Re: Problem starting master node
Date Wed, 16 May 2012 13:39:14 GMT
The problem isn't the hdfs layout, it's the zookeeper layout.

Use the zkCli.sh utility to look around:

$ zkCli.sh -server zoohost
prompt >  ls /accumulo/7f76ab57-ef3e-4750-**beda-401127a64cf0/masters

This location should have been created when the instance was created.

Also, I don't want to discourage you from using trunk, but we are making
lots of unstable changes to it.  Expect hiccups. If you want smooth
functionality, grab the 1.4 branch.

-Eric


On Wed, May 16, 2012 at 9:21 AM, Joe McMahon <joe.mcmahon@iswcorp.com>wrote:

> I'm running Accumulo 1.5.0-SNAPSHOT and Cloudera's hadoop-0.20.2-cdh3u4 on
> CentOS6.
>
> I init Accumulo fine but then when I run start-all.sh or start-here.sh, I
> get the following error
>
> 16 09:14:13,422 [security.SecurityUtil] INFO : Succesfully logged in as
> user accumulo/XXXXX
> 16 09:14:13,586 [server.Accumulo] INFO : Attempting to talk to zookeeper
> 16 09:14:13,707 [server.Accumulo] INFO : Zookeeper connected and
> initialized, attemping to talk to HDFS
> 16 09:14:13,708 [server.Accumulo] INFO : Connected to HDFS
> Thread "org.apache.accumulo.server.**master.state.SetGoalState" died null
> java.lang.reflect.**InvocationTargetException
>        at sun.reflect.**NativeMethodAccessorImpl.**invoke0(Native Method)
>        at
> sun.reflect.**NativeMethodAccessorImpl.**invoke(**
> NativeMethodAccessorImpl.java:**57)
>        at
> sun.reflect.**DelegatingMethodAccessorImpl.**invoke(**
> DelegatingMethodAccessorImpl.**java:43)
>        at java.lang.reflect.Method.**invoke(Method.java:616)
>        at org.apache.accumulo.start.**Main$1.run(Main.java:89)
>        at java.lang.Thread.run(Thread.**java:679)
> Caused by: org.apache.zookeeper.**KeeperException$**NoNodeException:
> KeeperErrorCode = NoNode for
> /accumulo/7f76ab57-ef3e-4750-**beda-401127a64cf0/masters/**goal_state
>        at
> org.apache.zookeeper.**KeeperException.create(**KeeperException.java:102)
>        at
> org.apache.zookeeper.**KeeperException.create(**KeeperException.java:42)
>        at org.apache.zookeeper.**ZooKeeper.create(ZooKeeper.**java:643)
>        at
> org.apache.accumulo.core.**zookeeper.ZooUtil.putData(**ZooUtil.java:146)
>        at
> org.apache.accumulo.core.**zookeeper.ZooUtil.**putPersistentData(ZooUtil.*
> *java:126)
>        at
> org.apache.accumulo.server.**zookeeper.ZooReaderWriter.**
> putPersistentData(**ZooReaderWriter.java:82)
>        at
> org.apache.accumulo.server.**master.state.SetGoalState.**
> main(SetGoalState.java:46)
>        ... 6 more
> Starting master on XXXXXX
> Starting garbage collector on XXXXXX
> Starting monitor on XXXXXX
> Starting tracer on XXXXXX
>
> When I look at HDFS, I see /accumulo is there but there is no directory as
> specified above in the error.  I have the following:
>
> drwxr-xr-x   - accumulo supergroup          0 2012-05-16 09:05
> /accumulo/instance_id
> drwxr-xr-x   - accumulo supergroup          0 2012-05-16 09:05
> /accumulo/tables
> drwxr-xr-x   - accumulo supergroup          0 2012-05-16 09:05
> /accumulo/version
>
> If I look at the instance_id directory, then I see an entry for that
> instance id:
>
> -sh-4.1$ hadoop fs -ls /accumulo/instance_id
> Found 1 items
> -rw-r--r--   2 accumulo supergroup          0 2012-05-16 09:05
> /accumulo/instance_id/**7f76ab57-ef3e-4750-beda-**401127a64cf0
>
> But it's not a directory.  Any ideas???
>
>

Mime
View raw message