Appreciate u replying. I have been poking around in Hive Github and I see you are a significant contributor to Druid integration 🙂 I set
<property>
<name>hive.druid.broker.address.default</name>
<value>10.93.206.102:8082</value>
</property>
in hive-site.xml and I am sure the datasource name I am using is correct because I can access Druid from Imply Pivot and Apache Superset using these params. Looking in the code for DruidStorageHandler I wonder if it may be something to do with me not setting the db type property anywhere?
final String dbType = HiveConf .getVar(SessionState.getSessionConf(), HiveConf.ConfVars.DRUID_METADATA_DB_TYPE);
as this would cause an exception in the constructor of the class?
if (dbType.equals("mysql")) { connector = new MySQLConnector(storageConnectorConfigSupplier, Suppliers.ofInstance(druidMetadataStorageTablesConfig) ); } else if (dbType.equals("postgresql")) { connector = new PostgreSQLConnector(storageConnectorConfigSupplier, Suppliers.ofInstance(druidMetadataStorageTablesConfig) ); } else { throw new IllegalStateException(String.format("Unknown metadata storage type [%s]", dbType)); }
I would have expected to see the exception logged at the hiveserver2 console on startup or when the [first] query is run... maybe? but I don't see anything but the semantic exception above.
