Please find below the link to the preliminary performance results for Messaging 1.0. For the impatiens, this is the summary:
* Messaging is better than JBossMQ in all categories we measured so far (for more details, see the graphs).
* In almost all cases JBossMQ couldn't keep up with Messaging at high send rate, choosing to crash instead. Messaging did not. This seems to be a database problem though, we will soon rerun the test with a real database, not HSQLDB.
* All graphs you see there are produced based on XML configurations and can be generated with only one ant command, so performance can be consistently tracked across releases. We see these "performance tests" as complements of the functional tests and integration tests. The goal is to automatically run these performance in the same way we do it with the functional tests.
* It is very easy to add another provider and build comparative graphs automatically. All it takes is to add to the test configuration file a section similar to:
<provider name="JBossMessaging">
<factory>org.jnp.interfaces.NamingContextFactory</factory>
<url>jnp://localhost:1099</url>
<pkg>org.jboss.naming:org.jnp.interfaces</pkg>
<executor name="REMOTE" url="rmi://localhost:7777/remote"/>
<executor name="REMOTE2" url="rmi://localhost:7777/remote2"/>
<executor name="LOCAL" url="rmi://localhost:7777/local-messaging"/>
<executor name="LOCAL2" url="rmi://localhost:7777/local-messaging2"/>
</provider>
The link to the performance results:
http://wiki.jboss.org/wiki/Wiki.jsp?page=JBossMessagingPerformanceResultsPre1_0 