[
https://issues.apache.org/jira/browse/CAMEL-26790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Work on CAMEL-26790 started by Claus Ibsen.
-------------------------------------------
> camel-chaos - an experiment that cannot tell does not pass: a fault that
> never fired, failures outside its expectations, an expectation of failures in
> a route with a dead letter channel
> -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: CAMEL-26790
> URL: https://issues.apache.org/jira/browse/CAMEL-26790
> Project: Camel
> Issue Type: Improvement
> Components: camel-jbang
> Reporter: Claus Ibsen
> Assignee: Claus Ibsen
> Priority: Major
> Fix For: 4.24.0
>
>
> Found in the recorded runs of {{camel cmd chaos-explore}} (CAMEL-26760) on
> the chaos example shop, read afterwards. Each made an experiment PASS that
> said nothing:
> # *A fault that never fired*: {{route=payment error rate=0.05}} fired 0 times
> over about 40 messages (chance, 12%), and the experiment passed. With no
> fault fired, the expectations say nothing about the fault: the verdict should
> be INCONCLUSIVE, with that reason.
> # *Failures outside the expectations*: an experiment expecting
> {{refunds:maxTime<=30s}} passed while 2 refunds failed. The experiment should
> list the routes that failed and that no expectation of it counts
> ({{failed}}), so neither the AI nor a person misses them.
> # *An expectation of failures in a route with a dead letter channel*:
> {{expect=parked:failed<=0}} on a route whose error handler is a dead letter
> channel is always met, as the dead letter channel handles what fails there (7
> orders that could not be parked were only logged). Adding such a rule should
> warn, as for a rule that matches nothing, and say that {{handled}} counts
> what the dead letter channel handled.
> The same pattern as the other fixes of these runs: Camel catches an
> experiment that cannot tell, and says why, so the model corrects itself.
> _Claude Code on behalf of Claus Ibsen_
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
