On Tue, Mar 8, 2011 at 4:21 PM, Loren J. Rittle <[email protected]> wrote:
> Hi Tom,
>
> I confess that I have been tearing down many clusters manually (since
> no .whirr/cluster directory is created when it hangs ;-) but I have
> been careful to completely remove all keys, groups and instances (via
> the AWS management console) that were created by whirr.
>
> If I remove zookeeper from the second set of machines, then I do not
> see the issue.  I tried the version posted multiple times (at least 3
> times across version .3 and .4) and have seen the error log posted
> each time.

I think this is the problem: you can't have a role split across two
template groups. In your configuration you have

whirr.instance-templates=1 zk+nn+jt+hbase-master,2 zk+dn+tt+hbase-regionserver

which splits the ZooKeeper ensemble across two groups. Until
https://issues.apache.org/jira/browse/WHIRR-249 is fixed, this won't
work.

As a workaround you can use more instances, e.g. by putting the
ZooKeeper ensemble on their own nodes. Alternatively you could use a
single node ZooKeeper cluster for a small cluster like this.

Cheers,
Tom

>
> Regards,
> Loren
>
> On Tue, Mar 8, 2011 at 5:49 PM, Tom White <[email protected]> wrote:
>> Hi Loren,
>>
>> I see this occasionally when a cluster hasn't been torn down properly
>> (e.g. the instances have been terminated manually but the EC2 security
>> groups are still there). I opened
>> https://issues.apache.org/jira/browse/WHIRR-249 for this case. Do you
>> think there may be other cases causing this?
>>
>> Thanks,
>> Tom
>>
>> On Tue, Mar 8, 2011 at 1:49 PM, Loren J. Rittle <[email protected]> 
>> wrote:
>>> Hello,
>>>
>>> While trying to start this cluster (pertaining portion of the properties 
>>> file):
>>>
>>> whirr.instance-templates=1 zk+nn+jt+hbase-master,2 
>>> zk+dn+tt+hbase-regionserver
>>> whirr.provider=ec2
>>>
>>> I received this message (which appears to hang the whirr
>>> launch-cluster process):
>>>
>>> Exception in thread "main" java.lang.IllegalStateException: The
>>> permission '50.16.131.205/32-1-2181-2181' has already been authorized
>>> on the specified group
>>>        at 
>>> org.jclouds.aws.handlers.ParseAWSErrorFromXmlContent.refineException(ParseAWSErrorFromXmlContent.java:106)
>>>        at 
>>> org.jclouds.aws.handlers.ParseAWSErrorFromXmlContent.handleError(ParseAWSErrorFromXmlContent.java:86)
>>>        at 
>>> org.jclouds.http.handlers.DelegatingErrorHandler.handleError(DelegatingErrorHandler.java:70)
>>>        at 
>>> org.jclouds.http.internal.BaseHttpCommandExecutorService$HttpResponseCallable.shouldContinue(BaseHttpCommandExecutorService.java:201)
>>>        at 
>>> org.jclouds.http.internal.BaseHttpCommandExecutorService$HttpResponseCallable.call(BaseHttpCommandExecutorService.java:166)
>>>        at 
>>> org.jclouds.http.internal.BaseHttpCommandExecutorService$HttpResponseCallable.call(BaseHttpCommandExecutorService.java:135)
>>>        at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
>>>        at java.util.concurrent.FutureTask.run(FutureTask.java:138)
>>>        at 
>>> java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:885)
>>>        at 
>>> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:907)
>>>        at java.lang.Thread.run(Thread.java:619)
>>> Caused by: org.jclouds.http.HttpResponseException: command: POST
>>> https://ec2.us-east-1.amazonaws.com/ HTTP/1.1 failed with response:
>>> HTTP/1.1 400 Bad Request; content: [The permission
>>> '50.16.131.205/32-1-2181-2181' has already been authorized on the
>>> specified group]
>>>        at 
>>> org.jclouds.aws.handlers.ParseAWSErrorFromXmlContent.handleError(ParseAWSErrorFromXmlContent.java:75)
>>>        ... 9 more
>>>
>>> I would be happy to provide more information if this is not a known issue.
>>>
>>> It seems that this exception should be caught and ignored.  Or,
>>> redundant requests should not be made in the first place.
>>>
>>> Regards,
>>> Loren
>>> --
>>> Loren J. Rittle, Principal Staff Engineer, Motorola Mobility (IL93)
>>> ljrittle@{motorola.com, acm.org, gcc.gnu.org, freebsd.org, gmail.com}
>>>
>>
>
>
>
> --
> Loren J. Rittle, Principal Staff Engineer, Motorola Mobility (IL93)
> ljrittle@{motorola.com, acm.org, gcc.gnu.org, freebsd.org, gmail.com}
>

Reply via email to