Hbase and Spark not starting in PredictionIO

[avinash v]

Hi All,

Can you please guide me how to solve this issue.
Actually i have installed predictionIo on top of cloudera distribution (5.3.2).
Ecosystem Versions in CDH are as follows:

1) Hbase (0.98.6-cdh5.3.2)2) spark (1.2.0-cdh5.3.2)3) elastic search (1.5.0)4) java jdk (1.7.0_67)5) predictionIo (0.9.1)6)

Hadoop (2.6.0 )

I have successfully installed the Elasticsearch and PredictionIO on CDH Cluster by following the below link:
http://docs.prediction.io/install/install-linux/

Now every thing is ready when i try to start the predictionIo i am getting the following error
Command to start the predictionIo is (./pio-start-all) and i am attaching the screenshot of the error which i am facing

Error:=====

$  ./pio-start-all

Starting Elasticsearch...Starting HBase..../pio-start-all: line 46: /opt/cloudera/parcels/CDH-5.3.2-

1.cdh5.3.2.p0.10/lib/hbase/bin/start-hbase.sh: No such file or directoryWaiting 10 seconds for HBase to fully

initialize...Starting PredictionIO Event Server...

After looking to this error i got to know that in hbase directory there is no file called start-hbase.sh. but there is another

file called start-hbase.cmd so i have tried to point to this file (start-hbase.cmd) i still get the same above  error.




Now i have also tried with another command called (pio status)

$ ./pio status
==========

here i am attaching another file related to this error.

I am also relating the error below:

Error:====

$ ./pio status
/usr/lib/PredictionIO-0.9.1/bin/semver.sh: line 12: =MAJOR_B: command not foundPredictionIO  Installed at:

/usr/lib/PredictionIO-0.9.1  Version: 0.9.1
Apache Spark  Installed at: /opt/cloudera/parcels/CDH-5.3.2-1.cdh5.3.2.p0.10/lib/sparkException in thread "main"

java.lang.ArrayIndexOutOfBoundsException: 1        at io.prediction.tools.console.Console$.status(Console.scala:949)        at

io.prediction.tools.console.Console$$anonfun$main$1.apply(Console.scala:615)        at io.prediction.tools.console.Console$

$anonfun$main$1.apply(Console.scala:577)        at scala.Option.map(Option.scala:145)        at

io.prediction.tools.console.Console$.main(Console.scala:577)        at io.prediction.tools.console.Console.main(Console.scala)

I got to know that there is some exception related to Spark. Here configurations related to Spark, Hbase and Elasticsearch are

default.
So, Please guide me how to rectify the above issues.


ThankYou in Advance.

Regards,
Avinash V

[Firman Gautama]

Hi Avinash,

try looking at https://groups.google.com/forum/#!topic/predictionio-user/DG_l14X_pY4, it may help you to get started. (CDH 5.4.1 HBase)

regards,

Firman