I tried making an hdfs storage plugin to the other cluster, and got an error from Drill Explorer:
Drill failed to execute the query: select * from INFORMATION_SCHEMA.SCHEMATA WHERE SCHEMA_NAME LIKE '%' ESCAPE '\' AND SCHEMA_NAME <> 'sys' AND SCHEMA_NAME <> 'INFORMATION_SCHEMA' [30027]Query execution error. Details:[ SYSTEM ERROR: IOException: Could not create FileClient My end goal is to see if I can run a join query between two tables. One table resides on hdfs in one cluster, and the other resides in maprfs on another cluster. Is something like this possible? Thanks, Scott On Wed, Aug 24, 2016 at 6:03 PM, Abhishek Girish <[email protected]> wrote: > Can you clarify what you are trying to accomplish? > > For query processing itself, Drillbits can only talk to each other within > the same Drill cluster. However, the data sources themselves can be on two > different clusters. For example, if you have data on DFS on a different > cluster, you can have your current Drillbit cluster access the same > remotely. It does not matter if the remote cluster have Drillbits > configured or not. And even if they do, they will not be involved in query > processing here. You only need to provide the connect string for the remote > cluster DFS in your storage plugin config. > > On Wed, Aug 24, 2016 at 10:51 AM, scott <[email protected]> wrote: > > > Hello, > > I'm wondering if it is possible to connect two Drill clusters? For > > instance, could I add a JDBC storage plugin in one cluster that connects > to > > another Drill cluster? > > > > Thanks, > > Scott > > >
