0
votes

I have set up a Hadoop single node cluster with pseudo distributed operations, and YARN running. I am able to use Spark JAVA API to run queries as a YARN-client. I wanted to go one step further and try Apache Drill on this "cluster". I installed Zookeeper that is running smoothly but I am not able to start drill and I get this log:

nohup: ignoring input

Error: Could not find or load main class org.apache.drill.exec.server.Drillbit

Any idea?

I am on Windows 10 with JDK 1.8.

1
What version of Drill you are using? How you try to start Drill? Do you want to run Drill under YARN? drill.apache.org/docs/launch-drill-under-yarn - Vitalii Diravka
I am using Drill 1.14. I start drill with the following command: drillbit.sh start I tried to follow the explanations for drill under YARN but I got a similar error when I run the command "drill-on-yarn.sh --site C:/BigData/drill/site start". > Could not find or load main class org.apache.drill.yarn.client.DrillOnYarn I suspect that there is a problem with my drill-install path. Here is how I setup the path in my drill-on-yarn.conf file: drill.yarn: { ... drill-install: { client-path: "C:\BigData\drill\apache-drill-1.14.0.tar.gz" } .... - DavidL

1 Answers

0
votes

DRILL CLASSPATH is not initialized in the process of running drillbit on your machine.
For the purpose to start Drill on Windows machine it is necessary to run sqlline.bat script, for example:

C:\bin\sqlline sqlline.bat –u "jdbc:drill:zk=local;schema=dfs"

See more info: https://drill.apache.org/docs/starting-drill-on-windows/