Could you make sure your zeppelin-env.sh have SPARK_HOME exported?

Zeppelin(0.6.0-SNAPSHOT) uses spark-submit command when SPARK_HOME is
defined, but your error shows that "please use spark-submit".

Thanks,
moon
On 2015년 10월 8일 (목) at 오후 9:14 Sourav Mazumder <[email protected]>
wrote:

> Hi Deepak/Moon,
>
> After seeing the stack trace of the error and the code
> org.apache.zeppelin.spark.SparkInterpreter.java I think this is surely a
> bug in Spark Interpreter code.
>
> The SparkInterpreter code is always calling the constructor of
> org.apache.spark.SparkContext to create a new Spark Context whenever the
> SparkInterpreter class is loaded by
> org.apache.zeppelin.interpreter.remote.RemoteInterpreterServer. And hence
> this error.
>
> I'm not sure whether the check for yarn-cluster is newly added in
> SparkContext.
>
> Attaching here the complete stack trace for your ease of reference.
>
> Regards,
> Sourav
>
> org.apache.spark.SparkException: Detected yarn-cluster mode, but isn't
> running on a cluster. Deployment to YARN is not supported directly by
> SparkContext. Please use spark-submit. at
> org.apache.spark.SparkContext.<init>(SparkContext.scala:378) at
> org.apache.zeppelin.spark.SparkInterpreter.createSparkContext(SparkInterpreter.java:339)
> at
> org.apache.zeppelin.spark.SparkInterpreter.getSparkContext(SparkInterpreter.java:149)
> at
> org.apache.zeppelin.spark.SparkInterpreter.open(SparkInterpreter.java:465)
> at
> org.apache.zeppelin.interpreter.ClassloaderInterpreter.open(ClassloaderInterpreter.java:74)
> at
> org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:68)
> at
> org.apache.zeppelin.interpreter.LazyOpenInterpreter.interpret(LazyOpenInterpreter.java:92)
> at
> org.apache.zeppelin.interpreter.remote.RemoteInterpreterServer$InterpretJob.jobRun(RemoteInterpreterServer.java:276)
> at org.apache.zeppelin.scheduler.Job.run(Job.java:170) at
> org.apache.zeppelin.scheduler.FIFOScheduler$1.run(FIFOScheduler.java:118)
> at java.util.concurrent.Executors$RunnableAdapter.call(Unknown Source) at
> java.util.concurrent.FutureTask.run(Unknown Source) at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(Unknown
> Source) at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(Unknown
> Source) at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown
> Source) at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown
> Source) at java.lang.Thread.run(Unknown Source)
>
> On Mon, Oct 5, 2015 at 12:57 PM, Sourav Mazumder <
> [email protected]> wrote:
>
>> I could execute following without any issue.
>>
>> spark-submit --class org.apache.spark.examples.SparkPi --master
>> yarn-cluster --num-executors 1 --driver-memory 512m --executor-memory 512m
>> --executor-cores 1 lib/spark-examples.jar 10
>>
>> Regards,
>> Sourav
>>
>> On Mon, Oct 5, 2015 at 12:04 PM, ÐΞ€ρ@Ҝ (๏̯͡๏) <[email protected]>
>> wrote:
>>
>>> did you try a test job with yarn-cluster (outside zeppelin) ?
>>>
>>> On Mon, Oct 5, 2015 at 11:48 AM, Sourav Mazumder <
>>> [email protected]> wrote:
>>>
>>>> Yes I have them setup appropriately.
>>>>
>>>> Where I am lost is I can see that interpreter is running spark-submit
>>>> but at some point of time it is switching to creating a spark context.
>>>>
>>>> May be, as you rightly mentioned, because of some permission issue it
>>>> is not able to run driver on yarn cluster. But what is that issue/required
>>>> configuration I'm not able to figure out.
>>>>
>>>> Regards,
>>>> Sourav
>>>>
>>>> On Mon, Oct 5, 2015 at 11:38 AM, ÐΞ€ρ@Ҝ (๏̯͡๏) <[email protected]>
>>>> wrote:
>>>>
>>>>> Do you have these settings configured in zeppelin-env.sh
>>>>>
>>>>> export JAVA_HOME=/usr/src/jdk1.7.0_79/
>>>>>
>>>>> export HADOOP_CONF_DIR=/etc/hadoop/conf
>>>>>
>>>>> Most likely you have this as your able to run with yarn-client.
>>>>>
>>>>>
>>>>> Looks like the issue is to not be able to run the driver program on
>>>>> cluster.
>>>>>
>>>>> On Mon, Oct 5, 2015 at 11:13 AM, Sourav Mazumder <
>>>>> [email protected]> wrote:
>>>>>
>>>>>> Yes. Spark is installed in the machine where zeppelin is running.
>>>>>>
>>>>>> The location of spark.yarn.jar is very similar to what you have. I'm
>>>>>> using IOP as distribution and it is the directory naming convention
>>>>>> specific to IOP which is different form hdp.
>>>>>>
>>>>>> And yes the setup works perfectly fine when I use master as
>>>>>> yarn-client and same setup for SPARK_HOME, HADOOP_CONF_DIR and
>>>>>> HADOOP_CLIENT>
>>>>>>
>>>>>> Regards,
>>>>>> Sourav
>>>>>>
>>>>>> On Mon, Oct 5, 2015 at 10:25 AM, ÐΞ€ρ@Ҝ (๏̯͡๏) <[email protected]>
>>>>>> wrote:
>>>>>>
>>>>>>> Is spark installed on your zeppelin machine ?
>>>>>>>
>>>>>>> I would to try these
>>>>>>>
>>>>>>> master yarn-client
>>>>>>> spark.home === SPARK INSTALLATION HOME directory on your zeppelin
>>>>>>> server.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Looking at  spark.yarn.jar , i see spark is installed at
>>>>>>> /usr/iop/current/spark-thriftserver/  . But why is it thirftserver
>>>>>>> (i do not know what is it).
>>>>>>>
>>>>>>> I have spark installed (unzip) on zeppelin machine at 
>>>>>>> /usr/hdp/2.3.1.0-2574/spark/spark/
>>>>>>>  (can be any location) and have spark.yarn.jar to
>>>>>>> /usr/hdp/2.3.1.0-2574/spark/spark/lib/spark-assembly-1.4.1-hadoop2.6.0.jar.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Oct 5, 2015 at 10:20 AM, Sourav Mazumder <
>>>>>>> [email protected]> wrote:
>>>>>>>
>>>>>>>> Hi Deepu,
>>>>>>>>
>>>>>>>> Here u go.
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>> Sourav
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> *Properties* name value args master yarn-cluster spark.app.name 
>>>>>>>> Zeppelin
>>>>>>>> spark.cores.max spark.executor.memory 512m spark.home spark.yarn.jar
>>>>>>>> /usr/iop/current/spark-thriftserver/lib/spark-assembly.jar 
>>>>>>>> zeppelin.dep.localrepo
>>>>>>>> local-repo zeppelin.pyspark.python python zeppelin.spark.concurrentSQL
>>>>>>>> false zeppelin.spark.maxResult 1000 zeppelin.spark.useHiveContext true
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, Oct 5, 2015 at 10:05 AM, ÐΞ€ρ@Ҝ (๏̯͡๏) <[email protected]
>>>>>>>> > wrote:
>>>>>>>>
>>>>>>>>> Can you share screen shot of your spark interpreter on zeppelin
>>>>>>>>> web interface.
>>>>>>>>>
>>>>>>>>> I have exact same deployment structure and it runs fine with right
>>>>>>>>> set of configurations.
>>>>>>>>>
>>>>>>>>> On Mon, Oct 5, 2015 at 7:56 AM, Sourav Mazumder <
>>>>>>>>> [email protected]> wrote:
>>>>>>>>>
>>>>>>>>>> Hi Moon,
>>>>>>>>>>
>>>>>>>>>> I'm using 0.6 SNAPSHOT which I built from latest git hub.
>>>>>>>>>>
>>>>>>>>>> I tried setting SPARK_HOME in zeppelin-env.sh. Also I could see
>>>>>>>>>> that the control goes to the appropriate IF-ELSE block in 
>>>>>>>>>> interpreter.sh by
>>>>>>>>>> putting some debug statement.
>>>>>>>>>>
>>>>>>>>>> But I get the same error as follows -
>>>>>>>>>>
>>>>>>>>>> org.apache.spark.SparkException: Detected yarn-cluster mode, but
>>>>>>>>>> isn't running on a cluster. Deployment to YARN is not supported 
>>>>>>>>>> directly by
>>>>>>>>>> SparkContext. Please use spark-submit. at
>>>>>>>>>> org.apache.spark.SparkContext.<init>(SparkContext.scala:378) at
>>>>>>>>>> org.apache.zeppelin.spark.SparkInterpreter.createSparkContext(SparkInterpreter.java:339)
>>>>>>>>>> at
>>>>>>>>>> org.apache.zeppelin.spark.SparkInterpreter.getSparkContext(SparkInterpreter.java:149)
>>>>>>>>>> at
>>>>>>>>>> org.apache.zeppelin.spark.SparkInterpreter.open(SparkInterpreter.java:465)
>>>>>>>>>> at
>>>>>>>>>> org.apache.zeppelin.interpreter.ClassloaderInterpreter.open(ClassloaderInterpreter.java:74)
>>>>>>>>>> at
>>>>>>>>>> org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:68)
>>>>>>>>>> at
>>>>>>>>>> org.apache.zeppelin.interpreter.LazyOpenInterpreter.interpret(LazyOpenInterpreter.java:92)
>>>>>>>>>> at
>>>>>>>>>> org.apache.zeppelin.interpreter.remote.RemoteInterpreterServer$InterpretJob.jobRun(RemoteInterpreterServer.java:276)
>>>>>>>>>> at org.apache.zeppelin.scheduler.Job.run(Job.java:170) at
>>>>>>>>>> org.apache.zeppelin.scheduler.FIFOScheduler$1.run(FIFOScheduler.java:118)
>>>>>>>>>> at 
>>>>>>>>>> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
>>>>>>>>>> at java.util.concurrent.FutureTask.run(FutureTask.java:266) at
>>>>>>>>>> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
>>>>>>>>>> at
>>>>>>>>>> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
>>>>>>>>>> at
>>>>>>>>>> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
>>>>>>>>>> at
>>>>>>>>>> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
>>>>>>>>>> at java.lang.Thread.run(Thread.java:745)
>>>>>>>>>>
>>>>>>>>>> Let me know if you need any other details to figure out what is
>>>>>>>>>> going on.
>>>>>>>>>>
>>>>>>>>>> Regards,
>>>>>>>>>> Sourav
>>>>>>>>>>
>>>>>>>>>> On Wed, Sep 30, 2015 at 1:53 AM, moon soo Lee <[email protected]>
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Which version of Zeppelin are you using?
>>>>>>>>>>>
>>>>>>>>>>> Master branch uses spark-submit command, when SPARK_HOME is
>>>>>>>>>>> defined in conf/zeppelin-env.sh
>>>>>>>>>>>
>>>>>>>>>>> If you're not on master branch, recommend try it with SPARK_HOME
>>>>>>>>>>> defined.
>>>>>>>>>>>
>>>>>>>>>>> Hope this helps,
>>>>>>>>>>> moon
>>>>>>>>>>>
>>>>>>>>>>> On Wed, Sep 23, 2015 at 10:21 PM Sourav Mazumder <
>>>>>>>>>>> [email protected]> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Hi,
>>>>>>>>>>>>
>>>>>>>>>>>> When I try to run Spark Interpreter in Yarn Cluster mode from a
>>>>>>>>>>>> remote machine I always get the error saying try spark-submit than 
>>>>>>>>>>>> using
>>>>>>>>>>>> spark context.
>>>>>>>>>>>>
>>>>>>>>>>>> Mu Zeppelin process runs in a separate machine remote to the
>>>>>>>>>>>> YARN cluster.
>>>>>>>>>>>>
>>>>>>>>>>>> Any idea why is this error ?
>>>>>>>>>>>>
>>>>>>>>>>>> Regards,
>>>>>>>>>>>> Sourav
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> Deepak
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Deepak
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Deepak
>>>>>
>>>>>
>>>>
>>>
>>>
>>> --
>>> Deepak
>>>
>>>
>>
>

Reply via email to