[
https://issues.apache.org/jira/browse/CAMEL-26421?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Work on CAMEL-26421 started by shashank.
----------------------------------------
> camel-core - a route called from a doTry does not use the error handler of
> its routing slip, multicast, recipient list or dynamic router: duplicate
> deliveries, no redelivery or a lost failure
> -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: CAMEL-26421
> URL: https://issues.apache.org/jira/browse/CAMEL-26421
> Project: Camel
> Issue Type: Bug
> Components: camel-core
> Reporter: shashank
> Assignee: shashank
> Priority: Major
>
> When a step of a {{doTry}} calls another route ({{direct}}, {{seda}} with
> {{waitForTaskToComplete}}, {{enrich}}, ...), the EIPs of the called route
> that wrap each output in the error handler (routing slip, dynamic router,
> multicast, recipient list) do not do so (main 017f4fd11c4a). The called route
> has its own error handler, and a plain {{to()}} in it is redelivered as
> usual, but these EIPs behave as if they were in a doTry of their own:
> * a routing slip redelivers the whole slip, so an endpoint that already
> succeeded gets the message again;
> * a multicast or recipient list does not redeliver the failed output (one
> attempt instead of 1 + maximumRedeliveries);
> * a dynamic router that counts its calls in an exchange property, as in the
> dynamic router documentation, loses the failure: the redelivery of the whole
> router "succeeds" because the bean returns null.
> The same happens with a doTry that has only a doFinally, and with a route
> called from the steps of a circuit breaker.
> {code:java}
> onException(IllegalStateException.class).maximumRedeliveries(2).redeliveryDelay(0);
> from("direct:r0").doTry().to("direct:r1").doCatch(Exception.class).to("mock:caught").end();
> from("direct:r1").routingSlip(constant("mock:m1,direct:fail"));
> from("direct:fail").errorHandler(noErrorHandler()).process(e -> { throw new
> IllegalStateException("boom"); });
> // without the doTry: mock:m1 gets 1 message, direct:fail 3 attempts
> // with it: mock:m1 gets 3 messages
> {code}
> *Cause*: {{TryProcessor}} sets the exchange property {{TryRouteBlock}} while
> the steps of the doTry run. {{RoutingSlip.createErrorHandler}} and
> {{MulticastProcessor.wrapInErrorHandler}} (multicast and recipient list) read
> it and leave out the error handler per output, because the steps of a doTry
> have no error handler (CAMEL-7275). The property stayed set while the called
> route ran: {{direct}} hands over the same exchange, {{seda}} and {{enrich}}
> copy it. A split is not affected (its steps are channels with the route's
> error handler).
> *Fix* (proposed): {{CamelInternalProcessor.RouteLifecycleAdvice}} removes
> {{TryRouteBlock}} when an exchange enters a route and sets the caller's value
> again when the route is done (as advice state, so recursion and asynchronous
> completion are covered). A doTry inside the called route still turns off the
> error handler per output of its own steps. Upgrade guide entry: the called
> route can make more delivery attempts than before; to keep the old behaviour,
> put the EIP in a doTry of the called route or use
> {{errorHandler(noErrorHandler())}} on it.
> *Tests*: {{DoTryCalledRouteErrorHandlerTest}} (camel-core, 18 tests): routing
> slip (also asynchronous), multicast, recipient list, dynamic router, doCatch,
> seda and enrich callers, recursion, a doTry inside the called route, the
> caller's scope after a successful and a failed call, and controls. 9 fail on
> main and pass with the fix. camel-core: no test class goes from passing to
> failing.
> *Related*: CAMEL-25114 restored the marker after a nested doTry; CAMEL-26332
> and CAMEL-26333 fix other defects of a route called from a doTry (other
> properties and classes); the fixes combine.
> *How it was found*: a Rocq specification of Camel's routing semantics,
> written from the documentation, compared with real Camel on generated routes,
> then reduced to a plain Java DSL test.
> *Affected*: main, 4.23.0 and 4.22.1 (same results).
> _Filed with Claude Code on behalf of allthingssecurity._
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
