hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Harsh J <ha...@cloudera.com>
Subject Re: Live nodes count not incremented with the addition of new nodes
Date Fri, 24 Aug 2012 11:49:22 GMT
Each of your DN needs its own dfs.data.dir directory (configurable in
hdfs-site.xml), unique from others. Otherwise, it may not work the way
you intend it to.

On Fri, Aug 24, 2012 at 2:27 PM, iwannaplay games
<funnlearnforkids@gmail.com> wrote:
> yeah installation is proper but it seems they are sharing the same storage id'
> Do i need to make any change in slaves configuration.I have copied the
> slaves image on other nodes.
> Check this log file:
> 2012-08-24 19:46:18,179 INFO org.apache.hadoop.net.NetworkTopology:
> Adding a new node: /default-rack/192.168.0.190:50010
> 2012-08-24 19:46:18,188 INFO org.apache.hadoop.hdfs.StateChange:
> *BLOCK* NameNode.blocksBeingWrittenReport: from 192.168.0.190:50010 0
> blocks
> 2012-08-24 19:46:18,215 INFO org.apache.hadoop.hdfs.StateChange:
> *BLOCK* NameSystem.processReport: from 192.168.0.190:50010, blocks: 0,
> processing time: 1 msecs
> 2012-08-24 19:46:18,276 INFO org.apache.hadoop.hdfs.StateChange:
> BLOCK* NameSystem.registerDatanode: node registration from
> 192.168.0.122:50010 storage
> DS-2050366683-192.168.0.122-50010-1344447057014
> 2012-08-24 19:46:18,276 INFO org.apache.hadoop.hdfs.StateChange:
> BLOCK* NameSystem.registerDatanode: node 192.168.0.190:50010 is
> replaced by 192.168.0.122:50010 with the same storageID
> DS-2050366683-192.168.0.122-50010-1344447057014
> 2012-08-24 19:46:18,276 INFO org.apache.hadoop.net.NetworkTopology:
> Removing a node: /default-rack/192.168.0.190:50010
> 2012-08-24 19:46:18,276 INFO org.apache.hadoop.net.NetworkTopology:
> Adding a new node: /default-rack/192.168.0.122:50010
>
>
> Why with another node registration previous node is removed?
>
> Regards
> Prabhjot
>
> On Thu, Aug 23, 2012 at 7:24 PM, Harsh J <harsh@cloudera.com> wrote:
>>
>> I'd first check if your additional nodes came up fine (check the
>> services' logs) and connected to the master alright.
>>
>> On Thu, Aug 23, 2012 at 7:18 PM, iwannaplay games
>> <funnlearnforkids@gmail.com> wrote:
>> >
>> >
>> > Hi
>> >
>> > I had a 2 node cluster earlier.I added 2 nodes on the fly in hadoop cluster.
>> >
>> > Ran these commands
>> > hadoop dfsadmin -refreshNodes
>> > hadoop mradmin -refreshNodes
>> >
>> > But still when i check through fsck or http://master:50070
>> > i get only 2 live nodes,one master and one slave(out of 3 slaves,randomly it
>> > picks anyone)
>> >
>> > Can anybody tell me why its not updating the count of live nodes to 4.
>> >
>> > Regards
>> > Prabhjot
>> >
>>
>>
>>
>> --
>> Harsh J



-- 
Harsh J

Mime
View raw message