Does RapidMiner close the DB session after it finishes execution on given SQL query?
Hi,
I would like to know if RapidMiner closes the session in database after it finishes the execution of SQL query? In my case I use Teradata driver. Of course while working I ran process several times. I use 1 box for reading database table and another box for writing into datbase. But unfortunately after several repetitions my process starts failing with error:
[SQLState HY000] TDWM Throttle violation for Concurrent Sessions: For Rule Name 'Limit Adhoc User connects', Limit of 8 concurrent requests
It is just annoying. I know the number of 8 concurrent sessions is not big. I work in corporate world, so increasing the number of concurrent session may be time consuming due to some politics. Besides all other tools we use close sessions after finishing the execution. I did not have to deal with this problem before.
I would like to know if RapidMiner closes the session in database after it finishes the execution of SQL query? In my case I use Teradata driver. Of course while working I ran process several times. I use 1 box for reading database table and another box for writing into datbase. But unfortunately after several repetitions my process starts failing with error:
[SQLState HY000] TDWM Throttle violation for Concurrent Sessions: For Rule Name 'Limit Adhoc User connects', Limit of 8 concurrent requests
It is just annoying. I know the number of 8 concurrent sessions is not big. I work in corporate world, so increasing the number of concurrent session may be time consuming due to some politics. Besides all other tools we use close sessions after finishing the execution. I did not have to deal with this problem before.
Find more posts tagged with
Sort by:
1 - 1 of
11
in my experience RapidMiner Studio disconnects from the database after the operator finishes. RapidMiner Server doesn't, as it uses a connection pool for better database performance.
Just tested with PostgreSQL (don't have a Teradata instance to access): The database logs the disconnections as expected, in a process that reads and then writes.
Are you sure you aren't using the built-in connection pooling with your Teradata JDBC driver?
https://developer.teradata.com/doc/connectivity/jdbc/reference/current/jdbcug_chapter_3.html#CIHHACFH
If you do, you can decrease the number of open connections.
Connection pooling would explain what you see, but it's probably not coming from Studio.
Regards,
Balázs