OpenTelemetryDestroyer flushing

[jimkont]

following up from #35686 (comment)

in

quarkus/extensions/opentelemetry/runtime/src/main/java/io/quarkus/opentelemetry/OpenTelemetryDestroyer.java

Lines 17 to 18 in 09cd879

	openTelemetrySdk.getSdkTracerProvider().forceFlush();
	openTelemetrySdk.getSdkTracerProvider().shutdown();

Both calls return a CompletableResultCode, which is effectively a CompletableFuture. Should we use the .join(SECONDS, TimeUnit.SECONDS) in both calls to ensure the thread completes?
In the context of an HTTP server, with very few shutdowns, this should have a minimal effect, but in AWS Lambda calls, where the shutdowns are very frequent, it could possibly make a difference.

also related to #29448

[jimkont]

moving to Q&A