If you were able to start Drill correctly in distributed mode, then it should
have registered with Zookeeper.
To confirm his, you can look at the Drillbit logs.
You should see something like this, indicating the port number (0.0.0.0:8047):
2017-08-09 22:54:22,253 kk127 [main] INFO org.eclipse.jetty.server.Server -
jetty-9.1.5.v20140505
2017-08-09 22:54:23,300 kk127 [main] INFO o.e.j.server.handler.ContextHandler
- Started o.e.j.s.ServletContextHandler@7ebcdf3{/,null,AVAILABLE}
2017-08-09 22:54:23,307 kk127 [main] INFO o.e.jetty.server.ServerConnector -
Started ServerConnector@13c8e042{HTTP/1.1}{0.0.0.0:8047}
2017-08-09 22:54:23,307 kk127 [main] INFO org.eclipse.jetty.server.Server -
Started @11616ms
2017-08-09 22:54:23,308 kk127 [main] INFO o.apache.drill.exec.server.Drillbit
- Startup completed (4863 ms).
As for the connection via the SQLLine command... do you have the ZKs running on
your "NameNode", "DataNode1", "DataNode2" servers?
The way the distributed mode works is that during startup, the Drillbits
register with the Zookeeper, while a JDBC client like SQLLine would look up the
list of Drillbits in ZK to connect to one of them.... or directly connect to
the Drillbit (default port in this case is 31010).
You'll need to figure out which of these are out of line in your settings.
-----Original Message-----
From: Divya Gehlot [mailto:[email protected]]
Sent: Thursday, August 10, 2017 7:03 PM
To: [email protected]
Subject: Re: set up Apache Drill on Windows server in distributed mode
Hi,
Even I used in embedded mode and able to set up successfully Keeping that in
mind try to set up in distributed in windows.
I am able to set up Zookeeper and it is running fine .
I started Drill thorugh drillbit.sh start command in git bash it started
successfully.
*Web Console Error :*
but when I try to open web console using http://<one of the ip or host name of
my 3 nodes>:8047 Its giving page cannot be displayed
*SqlLine Errors :*
1.Direct drill command sqlline.bat -u
"jdbc:drill:zk=NameNode:2181,DataNode1:2181,DataNode2:2181"
> Error :
> C:\apache-drill-1.11.0\bin>sqlline.bat -u
> "jdbc:drill:drillbit=NameNode:2181,Da
> taNode1:2181,DataNode2:2181"
> DRILL_ARGS - " -u
> jdbc:drill:drillbit=NameNode:2181,DataNode1:2181,DataNode2:218
> 1"
> HADOOP_HOME not detected...
> HBASE_HOME not detected...
> Calculating Drill classpath...
> Error: Failure in connecting to Drill:
> org.apache.drill.exec.rpc.RpcException: H ANDSHAKE_COMMUNICATION :
> Channel closed /<ip3>:64513 <--> DataNode1/10.20.
> 6.158:2181. (state=,code=0)
> java.sql.SQLException: Failure in connecting to Drill:
> org.apache.drill.exec.rpc
> .RpcException: HANDSHAKE_COMMUNICATION : Channel closed /<ip3>:64513
> <--> DataNode1/<ip2>2181.
> at
> org.apache.drill.jdbc.impl.DrillConnectionImpl.<init>(DrillConnection
> Impl.java:167)
> at
> org.apache.drill.jdbc.impl.DrillJdbc41Factory.newDrillConnection(Dril
> lJdbc41Factory.java:72)
> at
> org.apache.drill.jdbc.impl.DrillFactory.newConnection(DrillFactory.ja
> va:69)
> at
> org.apache.calcite.avatica.UnregisteredDriver.connect(UnregisteredDri
> ver.java:143)
> at org.apache.drill.jdbc.Driver.connect(Driver.java:72)
> at sqlline.DatabaseConnection.connect(DatabaseConnection.java:167)
> at
> sqlline.DatabaseConnection.getConnection(DatabaseConnection.java:213)
2. Sql line command using Zookeeper by specifying the IPs of the machines :
C:\apache-drill-1.11.0\bin>sqlline -u "jdbc:drill:zk=<ip1>:2181,<ip2>:
:2181<ip3>::2181"
DRILL_ARGS - " -u jdbc:drill:zk=<ip1>:2181,<ip2>:
:2181<ip3>::2181"
Calculating HADOOP_CLASSPATH ...
HBASE_HOME not detected...
Calculating Drill classpath...
ûu (The system cannot find the file specified)
jdbc:drill:zk=<ip1>:2181,<ip2>:2181<ip3>:2181 (The filename,
directory name, or volume label syntax is incorrect)
apache drill 1.11.0
"what ever the mind of man can conceive and believe, drill can query"
sqlline>
3.Strating sqlline by specifying the hostname :
C:\apache-drill-1.11.0\bin>sqlline.bat -u "jdbc:drill:zk=NameNode,DataNode1,Data
Node2"
DRILL_ARGS - " -u jdbc:drill:zk=NameNode,DataNode1,DataNode2"
HADOOP_HOME not detected...
HBASE_HOME not detected...
Calculating Drill classpath...
ûu (The system cannot find the file specified)
jdbc:drill:zk=NameNode,DataNode1,DataNode2 (The filename, directory name, or vol
ume label syntax is incorrect)
apache drill 1.11.0
"got drill?"
sqlline>
Note : Removed the actual IP for security purpose.
Appreciate the help .
Thanks,
Divya
On 11 August 2017 at 08:20, Kunal Khatua <[email protected]> wrote:
> Most people have used Apache Drill on Windows primarily in Embedded
> mode because no one appears to have tried for more than 1 Drillbit.
>
> That said, you should be able to run Apache Drill in a distributed
> mode as well, since it is Java-based and would not need to rely on
> anything more than a Zookeeper.
>
> -----Original Message-----
> From: Divya Gehlot [mailto:[email protected]]
> Sent: Thursday, August 10, 2017 1:45 AM
> To: [email protected]
> Subject: set up Apache Drill on Windows server in distributed mode
>
> Hi,
>
> Has anybody set up Apache Drill cluster in Windows machines/servers?
>
>
> Thanks,
> Divya
>
