I would expect that to work.  What exactly is the error?

On Wed, Aug 27, 2014 at 6:02 AM, Matt Chu <m...@kabam.com> wrote:

> (apologies for sending this twice, first via nabble; didn't realize it
> wouldn't get forwarded)
>
> Hey, I know it's not officially released yet, but I'm trying to understand
> (and run) the Thrift-based JDBC server, in order to enable remote JDBC
> access to our dev cluster.
>
> Before asking about details, is my understanding of this correct?
> `sbin/start-thriftserver` is a JDBC/Hive server that doesn't require
> running a Hive+MR cluster (i.e. just Spark/Spark+YARN)?
>
> Assuming yes, I have hope that it all basically works, just that some
> documentation needs to be cleaned up:
>
> - I found a release page implying that 1.1 will be released "pretty
> soon-ish": https://cwiki.apache.org/confluence/display/SPARK/Wiki+Homepage
> - I can find recent (more recent 30 days or so) activity with promising
> titles: ["Updated Spark SQL README to include the hive-thriftserver
> module"](https://github.com/apache/spark/pull/1867), ["[SPARK-2410][SQL]
> Merging Hive Thrift/JDBC server (with Maven profile fix)"](
> https://github.com/apache/spark/pull/1620)
>
> Am I following all the right email threads, issues trackers, and whatnot?
>
> Specifically, I tried:
>
> 1. Building off of `branch-1.1`, synced as of ~today (2014 Aug 25)
> 2. Running `sbin/start-thriftserver.sh` in `yarn-client` mode
> 3. Can see the processing running, and the spark context/app created in
> yarn logs,
> and can connect to the thrift server on the default port of 10000 using
> `bin/beeline`
> 4. However, when I try to find out what that cluster has via `show
> tables;`, in the logs
> I see a connection error to some (what I assume to be) random port.
>
> So what service am I forgetting/too ignorant to run? Or did I
> misunderstand and we do need a live Hive instance to back thriftserver? Or
> is this a YARN-specific issue?
>
> Only recently started learning the ecosystem and community, so apologies
> for the longer post and lots of questions. :)
>
> Matt
>

Reply via email to