Apache Hive 1.2.0 Metastore and Presto - Error occurred during processing of message.

[brenda...@gmail.com]

Hi,

I feel like I am doing something very basic wrong here but can't figure out what it might be.

I have setup Apache Hadoop 2.7.0 and Apache Hive 1.2.0 and Presto 0.105.
I run the following:

$HIVE_HOME/bin/hiveserver2
$HIVE_HOME/bin/beeline -u jdbc:hive2://localhost:10000

From here I am able to happily issue hive queries / statements.

I then configure a Presto catalog for Hive as follows:
connector.name=hive-hadoop2
hive.metastore.uri=thrift://localhost:10000

However when I now connect with Presto and issue a statement I get a socket timeout.
I can telnet to localhost 10000 fine and it looks like Presto can connect but Presto and Hive are not talking the same language (see Hive log below).

I notice that Beeline is using JDBC protocol but I want thrift, but Hive thrift port is set to 10000.

What am I doing wrong or what are some suggestions on next steps for debugging?

Thanks in advance
Brendan

Presto Error:
Query 20150601_052745_00003_tjmqv failed: localhost: java.net.SocketTimeoutException: Read timed out

HIVE LOG

15/06/01 15:33:20 [HiveServer2-Handler-Pool: Thread-32]: ERROR server.TThreadPoolServer: Error occurred during processing of message.
java.lang.RuntimeException: org.apache.thrift.transport.TTransportException: Invalid status -128
at org.apache.thrift.transport.TSaslServerTransport$Factory.getTransport(TSaslServerTransport.java:219)
at org.apache.thrift.server.TThreadPoolServer$WorkerProcess.run(TThreadPoolServer.java:268)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
Caused by: org.apache.thrift.transport.TTransportException: Invalid status -128
at org.apache.thrift.transport.TSaslTransport.sendAndThrowMessage(TSaslTransport.java:232)
at org.apache.thrift.transport.TSaslTransport.receiveSaslMessage(TSaslTransport.java:184)
at org.apache.thrift.transport.TSaslServerTransport.handleSaslStartMessage(TSaslServerTransport.java:125)
at org.apache.thrift.transport.TSaslTransport.open(TSaslTransport.java:271)
at org.apache.thrift.transport.TSaslServerTransport.open(TSaslServerTransport.java:41)
at org.apache.thrift.transport.TSaslServerTransport$Factory.getTransport(TSaslServerTransport.java:216)
... 4 more

[David Phillips]

There are two different services which both use Thrift as the protocol: metastore and hiveserver2. Presto talks to the metastore which by default uses port 9083. The Hive JDBC driver uses hiveserver2.

> --
> You received this message because you are subscribed to the Google Groups "Presto" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to presto-users...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[brenda...@gmail.com]

Ok thanks David, figured something like this was the problem.
I'm really going to show my ignorance now...

How would one start the Apache Hive metastore?
I've been looking but I always seem to be stumbling on the command to start the Hive server (that runs the JDBC interface).

When I run this I can't see anything listening on port 9083 via netstat :(

[brenda...@gmail.com]

Ok, worked it out. Not sure how I missed it before because I am pretty sure I read this page.

For prosperity the command is:
$ hive --service metastore

Documented here - https://cwiki.apache.org/confluence/display/Hive/AdminManual+MetastoreAdmin

Thanks David for your help!

[Stephen Sprague]

good man.  he ground it out on his own! :)

[Damien Carol]

Why prefer release 2.7.0 over stable 2.6.0?