Hi James,
It looks like by doing "routeType.addInterceptor(processor)", it
results in wrapping each node(s) with the referenced processor. If I
have a route like: from("direct:start).to("mock:end1", "mock:end2"),
it ends up having an instrumentation processor wrapping "sendTo
mock:end1" and "sendTo mock:end2" processors. But, the referenced
processor is supposed (in my mind) to be a "per route" interceptor
that records exchanges processed by the entire route. So, a single
exchange sent to the route could cause the per-route instrumentation
processor to record two processed exchanges.
InstrumentationInterceptStrategy already intercepts each node to
record per-processor performance data. So, we probably don't need to
wrap each node again. Perhaps, I am missing something.
Thanks,
William
---
activemq/camel/trunk/camel-core/src/main/java/org/apache/camel/management/InstrumentationLifecycleStrategy.java
(original)
+++
activemq/camel/trunk/camel-core/src/main/java/org/apache/camel/management/InstrumentationLifecycleStrategy.java
Tue May 20 06:25:00 2008
@@ -145,13 +145,14 @@
ProcessorType<?>[] outputs =
routeType.getOutputs().toArray(new ProcessorType<?>[0]);
- routeType.clearOutput();
+ //routeType.clearOutput();
InstrumentationProcessor processor = new InstrumentationProcessor();
- routeType.intercept(processor);
- for (ProcessorType<?> output : outputs) {
+ routeType.addInterceptor(processor);
+
+ /*for (ProcessorType<?> output : outputs) {
routeType.addOutput(output);
}
-
+*/
interceptorMap.put(endpoint, processor);
}
}
