ExtremeTransactions 5.0.94

182106: Improve TransactionService shutdown: coordinator dispose should wait for active timer threads

We improved shutdown to wait for active timer threads, so there are less warnings in Tomcat installations (concerning pending threads and possible memory leaks).

182134: Make localTransactionMode = true by default (for JDBC)

We've reverted a decision we made when releasing the 5.0: localTransactionMode should not be false (by default) but rather true. This is to avoid backward compatibility issues with existing installations that want to upgrade.

182161: Pool problem with Shadow JDBC driver

We fixed an incompatibility in the way our pooled connections are initialised and a Shadow JDBC driver limitation.

182138: RestTransactionService: improved recovery

We improved distributed recovery for com.atomikos.remoting.taas.RestTransactionServiceImp.

182132: DefaultExportingTransactionManager: automatically add URL of recovery coordinator to Propagation

We have improved remoting recovery by automatically adding the URL of the recovery coordinator. Previously this required setting a separate manual startup property.

181667 / 181668: Improved how remoting recovery deals with timeouts

Abandoned or heuristic transactions should not delete the remote participant information - so remoting recovery can still do its job.

182099: JMS: createConnection(user, password) should log a warning

We previously lowered the log to DEBUG. For customers that don't want a warning, please:

1. Subclass AtomikosConnectionFactoryBean
2. Override this method to not log a warning

182041: Add DEBUG logs to BatchingMessageListenerContainer

We added debug logs to make it easier to see if batching is working.

181866: Improve remoting recovery to avoid remote calls if there is a shared LogCloud DB

When a shared DB is used, all information is present in the DB itself - so remote http calls are not needed. This has been implemented for recovery of imported transactions.

Only for customers. Not a customer yet?