Hi Andrei,
this cluster is still running, I am running a distcp job to copy my
data from S3 to HDFS.

The NameNode (via the Web UI) is sitll reporting:

Live Nodes      :       8
Dead Nodes      :       0
Decommissioning Nodes   :       0

I do not see errors in the logs.

I can try to connect to one of the machines which did not join the cluster,
but I am not sure what to do to make it join the cluster once I am connected
to it.

Paolo

On 2 November 2011 15:29, Andrei Savu <[email protected]> wrote:
> Are you seeing any errors in the logs? Can you check one of the machines
> that failed to join the cluster?
> Are you sure they've tried to join the rest of the cluster? Maybe you have
> to wait a bit more.
>
> -- Andrei Savu
>
> On Wed, Nov 2, 2011 at 5:25 PM, Paolo Castagna
> <[email protected]> wrote:
>>
>> Hi
>>
>> On 2 November 2011 14:59, Paolo Castagna <[email protected]>
>> wrote:
>> > Hi Andrei,
>> > I've just tried again, the only difference in the recipe:
>> > whirr.instance-templates=1 hadoop-namenode+hadoop-jobtracker,16
>> > hadoop-datanode+hadoop-tasktracker
>> > I saw the same exception, but now I can connect to the web UIs as usual.
>>
>> Well, I spoken too soon.
>>
>> The very same cluster had 17 instances, I can see all of them running
>> via the Amazon console (i.e. I am paying for them), however the
>> NameNode and the JobTracker see only 8 nodes. :-(
>>
>> Paolo
>
>

Reply via email to