I'm moving some code from a proprietary connection pool to a pooled DataSource in WebSphere 5.x. The authors of the existing pool put a lot of work into error recovery. One strategy is that when a SQL exception indicates communication errors they purge the pool and let it restart from scratch, which neatly recovers from a brief network burp.
I'm having trouble finding Sun doc on what to expect from a DataSource in such a situation. There's no API to purge the pool, so I'm hoping it has some signficant smarts built in. Any pointers to doc on this?
We'll be on UDB and I hope we'll be able to use Automatic Client Reroute. I believe that was a sufficient solution on another system in the company. It's not clear if I can use ACR with a single db instance, not in an HADR cluster. Anybody know for sure?
A good question is never answered. It is not a bolt to be tightened into place but a seed to be planted and to bear more seed toward the hope of greening the landscape of the idea. John Ciardi
Joined: Jan 29, 2003
Ok, I'm told my old team uses ACR on databases without HADR. Whew.
Still don't know what IBM or Sun promises for DataSources reconnecting after transient network burp or a major stop & restart of the database.