Re: increasing Spark driver memory using Zeppelin with Spark/YARN

2016-02-03 Thread Hyung Sung Shim
Hello. I commented below your questions. 1. is the Spark driver executing in the same JVM as the Zeppelin RemoteInterpreterServer? *-> I know that is sperated. The interpreter server process is executed by ~/bin/interpreter.sh* 2. how does one correctly size the Spark driver memory (Xmx) in

Re: zeppelin multi user mode?

2016-02-03 Thread Hyung Sung Shim
Hello yunfeng. You can also refer to https://github.com/NFLabs/z-manager/tree/master/multitenancy. Thanks. 2016-02-04 3:56 GMT+09:00 Christopher Matta : > I have had luck with a single Zepplin installation and config directories > in each user home directory. That way each

Re: zeppelin multi user mode?

2016-02-03 Thread Benjamin Kim
I forgot to mention that I don’t see Spark 1.6 in the list of versions when installing z-manager. > On Feb 3, 2016, at 10:08 PM, Corneau Damien wrote: > > @Benjamin, > We do support version 1.6 of Spark, see: >

Re: HBase Interpreter

2016-02-03 Thread Rajat Venkatesh
Can you check the version of HBase ? HBase interpreter has been tested with HBase 1.0.x and Hadoop 2.6.0. There is a good chance this error is due to mismatch in versions. On Thu, Feb 4, 2016 at 10:20 AM Benjamin Kim wrote: > I got this error below trying out the new HBase

Re: HBase Interpreter

2016-02-03 Thread Rajat Venkatesh
Oh. That should work. I've tested with 1.0.0. Hmm On Thu, Feb 4, 2016 at 10:50 AM Benjamin Kim wrote: > Hi Rajat, > > The version of HBase that comes with CDH 5.4.8 is 1.0.0. How do I check if > they are compatible? > > Thanks, > Ben > > > On Feb 3, 2016, at 9:16 PM, Rajat

Re: Upgrade spark to 1.6.0

2016-02-03 Thread Felix Cheung
I think his build command only works with Cloudera CDH 5.4.8, as you can see. Mismatch Akka version is very common if the Hadoop distribution is different. What version of Spark and Hadoop distribution are you running with? On Tue, Feb 2, 2016 at 1:36 PM -0800, "Daniel Valdivia"

Re: Upgrade spark to 1.6.0

2016-02-03 Thread Daniel Valdivia
I don't need any specific version of Hadoop, I actually removed it from the build command and still get the error, I just need spark 1.6 > On Feb 3, 2016, at 9:05 AM, Felix Cheung wrote: > > I think his build command only works with Cloudera CDH 5.4.8, as you can

Re: zeppelin multi user mode?

2016-02-03 Thread Christopher Matta
I have had luck with a single Zepplin installation and config directories in each user home directory. That way each user gets their own instance and will not interfere with each other. You can start the Zepplin server with a config flag pointing to the config directory. Simply copy the config