Github user exabrial commented on a diff in the pull request:
https://github.com/apache/tomee/pull/175#discussion_r223483193
--- Diff:
container/openejb-jpa-integration/src/main/java/org/apache/openejb/jpa/integration/eclipselink/OpenEJBServerPlatform.java
---
@@ -26,13 +27,14 @@
import javax.transaction.Synchronization;
import javax.transaction.TransactionManager;
-public class OpenEJBServerPlatform extends JMXServerPlatformBase {
+public class OpenEJBServerPlatform extends JMXServerPlatformBase
implements JMXEnabledPlatform {
public OpenEJBServerPlatform(final DatabaseSession newDatabaseSession)
{
super(newDatabaseSession);
try {
mBeanServer = MBeanServer.class.cast(
OpenEJBServerPlatform.class.getClassLoader().loadClass("org.apache.openejb.monitoring.LocalMBeanServer")
.getMethod("get").invoke(null));
+ this.prepareServerSpecificServicesMBean();
} catch (final Exception e) {
// no-op
--- End diff --
I agree with you; I don't understand why they split it out either, but they
literally did it for _every other server_ so there's probably a good reason.
On the fail vs noop, that was there before me. I suggest leaving it so we
don't change current behavior (if JMX fails to initialize, we still alow the
application to run. Right now JMX is failing.)
> 6e2a4f7c419 (Thiago Veronezi 2015-11-23 14:38:43 -0500 36)
.getMethod("get").invoke(null));
> 515306f7ba0 (Jonathan S. Fisher 2018-09-29 13:10:49 -0500 37)
this.prepareServerSpecificServicesMBean();
> 6e2a4f7c419 (Thiago Veronezi 2015-11-23 14:38:43 -0500 38) }
catch (final Exception e) {
> 6e2a4f7c419 (Thiago Veronezi 2015-11-23 14:38:43 -0500 39)
// no-op
> 6e2a4f7c419 (Thiago Veronezi 2015-11-23 14:38:43 -0500 40) }
> 6e2a4f7c419 (Thiago Veronezi 2015-11-23 14:38:43 -0500 41) }
> 6e2a4f7c419 (Thiago Veronezi 2015-11-23 14:38:43 -0500 42)
---
