We are currently using dbcp 1.1 with oracle 10g but we're moving to an Oracle 11g RAC setup.
Our existing code doesn't have any "retry" logic in response to SQLExceptions so does dbcp subscribe to ONS events and what is the general approach it uses to become aware of a node down event in the cluster and how does it handle failover? ie. what connections are recycled in response to a node down event and what does the jdbc url have to look like if this is to take place? Our existing jdbc url currently references a virtual ip address which is frontending the cluster which I believe means that you won't have to wait for the TCP timeout when a node goes down but rather the onus is still on the client to catch the SQLException and have retry logic in order to reconnect to the new node. .... --------------------------------------------------------------------- To unsubscribe, e-mail: user-unsubscr...@commons.apache.org For additional commands, e-mail: user-h...@commons.apache.org