Standardize sender shutdown implementations

[jack-berg]

Let's hold off on this. I just did an analysis of the shutdown of OkHttpHttpSender, OkHttpGrpcSender, and JdkHttpSender, and they're all different. Will open a separate issue to standardize (probably mirroring the logic in OkHttpGrpcSender which is the newest).

Originally posted by @jack-berg in #8276

opentelemetry-java/exporters/sender/okhttp/src/main/java/io/opentelemetry/exporter/sender/okhttp/internal/OkHttpGrpcSender.java

Lines 301 to 337 in 7fcacb0

	public CompletableResultCode shutdown() {
	client.dispatcher().cancelAll();
	client.connectionPool().evictAll();
	if (managedExecutor) {
	ExecutorService executorService = client.dispatcher().executorService();
	// Use shutdownNow() to interrupt idle threads immediately since we've cancelled all work
	executorService.shutdownNow();
	// Wait for threads to terminate in a background thread
	CompletableResultCode result = new CompletableResultCode();
	Thread terminationThread =
	new Thread(
	() -> {
	try {
	// Wait up to 5 seconds for threads to terminate
	// Even if timeout occurs, we succeed since these are daemon threads
	boolean terminated = executorService.awaitTermination(5, TimeUnit.SECONDS);
	if (!terminated) {
	logger.log(
	Level.WARNING,
	"Executor did not terminate within 5 seconds, proceeding with shutdown since threads are daemon threads.");
	}
	} catch (InterruptedException e) {
	Thread.currentThread().interrupt();
	} finally {
	result.succeed();
	}
	},
	"okhttp-shutdown");
	terminationThread.setDaemon(true);
	terminationThread.start();
	return result;
	}
	return CompletableResultCode.ofSuccess();
	}

opentelemetry-java/exporters/sender/okhttp/src/main/java/io/opentelemetry/exporter/sender/okhttp/internal/OkHttpHttpSender.java

Lines 204 to 212 in 7fcacb0

	@Override
	public CompletableResultCode shutdown() {
	client.dispatcher().cancelAll();
	if (managedExecutor) {
	client.dispatcher().executorService().shutdownNow();
	}
	client.connectionPool().evictAll();
	return CompletableResultCode.ofSuccess();
	}

opentelemetry-java/exporters/sender/jdk/src/main/java/io/opentelemetry/exporter/sender/jdk/internal/JdkHttpSender.java

Lines 410 to 423 in 7fcacb0

	@Override
	public CompletableResultCode shutdown() {
	if (managedExecutor) {
	executorService.shutdown();
	}
	if (AutoCloseable.class.isInstance(client)) {
	try {
	AutoCloseable.class.cast(client).close();
	} catch (Exception e) {
	return CompletableResultCode.ofExceptionalFailure(e);
	}
	}
	return CompletableResultCode.ofSuccess();
	}

opentelemetry-java/exporters/sender/grpc-managed-channel/src/main/java/io/opentelemetry/exporter/sender/grpc/managedchannel/internal/UpstreamGrpcSender.java

Lines 231 to 237 in 7fcacb0

	@Override
	public CompletableResultCode shutdown() {
	if (shutdownChannel) {
	channel.shutdownNow();
	}
	return CompletableResultCode.ofSuccess();
	}

[ADITYA-CODE-SOURCE]

@jack-berg I'd be interested in working on this. The scope seems clear — standardize the shutdown logic of OkHttpHttpSender, JdkHttpSender, and UpstreamGrpcSender to follow the OkHttpGrpcSender pattern (cancel work, evict connection pool, wait for threads on a background daemon thread).
Is there anything specific I should watch out for, or any edge cases from the parent PR #8276 that should be carried over?