Binder doesn't respect a kill request until retries exhausted

[Mrc0113]

The solace cloud stream binder doesn't seem to respect a kill request prior to being connected. This is not ideal when developing in an IDE (This happens in eclipse and spring tool suite. Have not verified with Intellij or others). If you use the wrong credentials you can't hit the stop button until your retries are exhausted which can take quite some time since I think the default per host is 21.

To reproduce: Set the following properties and include an invalid password. Start your app via your IDE and watch it fail to connect forever even after you try to stop it

spring.cloud.stream.binders.solace.environment.solace.java.connectRetries=-1
spring.cloud.stream.binders.solace.environment.solace.java.connectRetriesPerHost=-1

[Mrc0113]

More info on this. The issue isn't not responding to a traditional kill request. A traditional kill request such as "ctrl+c" or killing it when running as a "java application" in eclipse work. However, the app doesn't respect a "inMXBeanRegistrar$SpringApplicationAdmin : Application shutdown requested." which is what Spring Tool Suite issues when you stop a Spring Boot app (started using "Run as" -> "Spring Boot App" and I believe is the proper request to shutdown Spring Boot apps.

I have not yet had the chance to verify if this is a framework issue or a binder issue by trying with a different binder.

[PhilippeKhalife]

Leaving in current state.