A connection pool is a set of created connections to database servers stored in cache. A connection pool stores a permanent set of opened connections and reuse them on multiuser work in a repository.
On requesting data, BI server creates a DBMS server connection and sends a request to retrieve data. The obtained response is sent to the web application where the user is working; the opened DBMS connection does not close and is moved to the connection pool. On next data requests, the requested connection is searched in the pool and is used, if possible.
By default connection pool is enabled and the maximum number of connections equal to 50 is set. To disable it or to limit the number of connections, use the MaxConnections parameter in the Pool subsection in registry or in the Settings.xml file.
IMPORTANT. If logic of a developed web application supposes multiuser work in one repository, it is mandatory to use a connection pool.
See also: