I have set up a cluster in Ambari.
I am using Aqua Data Studio to connect hbase using "org.apache.phoenix.jdbc.PhoenixDriver" drive to connect through ZK nodes.
I can run normal queries on hbase tables, but it throws an error while running any aggregate function like max()/count()
>[Error] Script lines: 6-10 ------------------------- org.apache.phoenix.exception.PhoenixIOException: The system cannot find the path specified [Executed: 23/12/2016 11:47:19 AM] [Execution: 28ms]
What I need to do to fix this issue?
What is the path set for "phoenix.spool.directory" ?
Also can you please stare the complete stackTrace?
Also on your Aqua data studio side do you have the proper client configuration files placed in the classpath? (like
I just checked from Ambari Metrics/Configs/Metric Collector/Advanced ams-hbase-site/phoenix.spool.directory
In Aqua Data Stufio, I have the jar files: hbase-client.jar;phoenix-client.jar;hbase-client-184.108.40.206.4.2.12-1.jar;phoenix-4.9.0-HBase-1.2-client.jar
- Can you please share the complete stackTrace?
- Also the Classpath setting that you shared contains Semi Colon separated values which indicates that you might be using Windows Environment on the client side, So i will suggest you to please go through the following article which talks about a similar issue on Windows:
"I just checked from Ambari Metrics/Configs/Metric Collector/Advanced ams-hbase-site/phoenix.spool.directory" are you trying to connect to the AMS HBase/Phoenix instance? Not the HBase service?
What about the value of hbase.tmp.dir? That property value would be substituted in phoenix.spool.directory. It defaults to "/tmp" if you do not have it set. This directory must exist and be writable by the user running the Phoenix queries.