Try to stabilize ZipkinWebClientSenderTests

[wilkinsona]

Several tests are flaky on CI. We should try to stabilize them.

[wilkinsona]

This hasn't helped:

reactor.core.Exceptions$ReactiveException: java.util.concurrent.TimeoutException: Did not observe any item or terminal signal within 60000ms in 'flatMap' (and no fallback has been configured)	
	at reactor.core.Exceptions.propagate(Exceptions.java:410)	
	at reactor.core.publisher.BlockingSingleSubscriber.blockingGet(BlockingSingleSubscriber.java:102)	
	at reactor.core.publisher.Mono.block(Mono.java:1779)	
	at org.springframework.boot.actuate.autoconfigure.tracing.zipkin.ZipkinWebClientSender$WebClientHttpPostCall.doExecute(ZipkinWebClientSender.java:76)	
	at org.springframework.boot.actuate.autoconfigure.tracing.zipkin.ZipkinWebClientSender$WebClientHttpPostCall.doExecute(ZipkinWebClientSender.java:54)	
	at zipkin2.Call$Base.execute(Call.java:391)	
	at org.springframework.boot.actuate.autoconfigure.tracing.zipkin.HttpSender.check(HttpSender.java:72)	
	at org.springframework.boot.actuate.autoconfigure.tracing.zipkin.ZipkinWebClientSenderTests.checkShouldNotRaiseException(ZipkinWebClientSenderTests.java:104)

I'm going to revert that change as, when the problem occurs, the build now takes longer as it waits for 60 seconds.

Looking more closely at Develocity's test dashboard, the flakiness appears to be limited to 3.2.x and way more common on Windows than Linux. Of the 43 flaky test runs in the last 28 days, 42 of them have occurred on Windows and 1 on Linux.

[wilkinsona]

The latest results show the problem occurring on 3.2.x, 3.3.x, and main with a more even split between Windows and Linux. Unfortunately, I can't reproduce locally and haven't found a way to address the flakiness on CI. I think we'll have to live with this for now.