Hi moon,
I see your point that there would be overhead in managing two systems. However,
I don’t believe that working within JIRA will achieve what I’m thinking of. I’m
impressed there are people who use JIRA and seem to be end users; however, I
speculate that these are advanced users – edging
Hi guys!
I find the suggestion to vote via trello totally cool and would support it.
So if everyone is OK with this, let's do this.
I was looking for such a possibility to have a community process to
prioritize something for quite some time (have also played with various
JIRA workarounds) - but
Hello !
I am trying to launch some very greedy processes on a Spark 1.4 Cluster using
Zeppelin, and I don't understand how to configure Spark memory properly. I’ve
tried to set SPARK_MASTER_MEMORY, SPARK_WORKER_MEMORY and SPARK_EXECUTOR_MEMORY
environment variables on the spark cluster nodes,
I am trying the simple thing in pyspark:
%pyspark
rdd = sc.parallelize([1,2,3])
print(rdd.collect())
z.show(sqlContext.createDataFrame(rdd))
AND keep getting error:
Traceback (most recent call last): File /tmp/zeppelin_pyspark.py, line
116, in module eval(compiledCode) File string, line 3, in
It looks like pyspark is not able to talk to its java classes. Could you double
check the SPARK_HOME etc are correctly set?
On Wed, Jul 22, 2015 at 1:03 PM -0700, Renxia Wang renxia.w...@gmail.com
wrote:
Hi guys,
I am trying to run zeppelin locally, interact with spark in local mode.
I ran
Hi,
Although SPARK_HOME correctly set and build Z with -Dpyspark I still experience
local class not compatible error when I running any pyspark in Z. Job was
correctly send and processed on cluster but the error seems thrown after final
stage (Python RDD deserialization?) I followed instruction
Guys,
thank you for great suggestions!
Am I right that you suggest using Trello not instead of ASF hosted
JIRA, but together with it, and are volunteering to support it as a
tool for prioritizing user's feedback?
Also, how do you think, should we then move further discussion to the
Hi,
thank you for your interest in Zeppelin!
You just have to set the 'Spark' interpreter properties in 'Interpreters' menu:
CPU
spark.cores.max: 24
Mem
spark.executor.memory 22g
You actually can use any of the
http://spark.apache.org/docs/latest/configuration.html#application-properties
