Hi Anil,
If you use derby as metastore where ever you try to start hive cli it will
create a metastore in that directory.
First option-Use other rdbms such as postgresql, mysql, oracle ...
Second option- conf.set("javax.jdo.option.ConnectionURL",
"jdbc:derby:;databaseName=/tmp/metastore_db;create=*false*"); so that only
one metastore will be there and only one user can connect at any time. If
another user want to connect your all sessions should be closed.
Note ( I dont know java but I could see you are using derby and we can
check these options)
Thanks,
Abirami
On Fri, Jul 11, 2014 at 9:25 AM, AnilKumar B <akumarb2...@gmail.com> wrote:
> Hi,
>
> I am trying to generate hive plan as below. But even after creating the
> "src" table, I am facing, Table not found Exception due to MetaStore issue.
> Can any one help me in resolving this?
>
>
> private Driver createDriver() {
> HiveConf conf = new HiveConf(Driver.class);
> conf.set("hive.metastore.warehouse.dir", "file:///tmp/warehouse");
> conf.set("javax.jdo.option.ConnectionURL",
> "jdbc:derby:;databaseName=/tmp/metastore_db;create=true");
> SessionState.start(conf);
> Driver driver = new Driver(conf);
> driver.init();
> return driver;
> }
>
> Driver driver = createDriver();
> driver.compile("create table src (c1 string, c2 string)");
> int rc = driver.compile("EXPLAIN LOGICAL SELECT key, count(1) FROM
> (SELECT key, value FROM src) subq1 GROUP BY key");
>
> if (rc != 0) {
> return;
> }
>
> QueryPlan plan = driver.getPlan();
> org.apache.hadoop.hive.ql.exec.Task<? extends Serializable> task = plan
> .getRootTasks().get(0);
> ExplainWork work = (ExplainWork) task.getWork();
>
>
>
> Thanks & Regards,
> B Anil Kumar.
>
