聊聊flink的slot.idle.timeout配置

序

本文主要研究一下flink的slot.idle.timeout配置

JobManagerOptions

flink-release-1.7.2/flink-core/src/main/java/org/apache/flink/configuration/JobManagerOptions.java

@PublicEvolving
public class JobManagerOptions {
	//......

	/**
	 * The timeout in milliseconds for a idle slot in Slot Pool.
	 */
	public static final ConfigOption<Long> SLOT_IDLE_TIMEOUT =
		key("slot.idle.timeout")
			// default matches heartbeat.timeout so that sticky allocation is not lost on timeouts for local recovery
			.defaultValue(HeartbeatManagerOptions.HEARTBEAT_TIMEOUT.defaultValue())
			.withDescription("The timeout in milliseconds for a idle slot in Slot Pool.");

	//......
}
复制代码

• slot.idle.timeout默认为HeartbeatManagerOptions.HEARTBEAT_TIMEOUT.defaultValue()，即50000L毫秒

SlotPool

flink-release-1.7.2/flink-runtime/src/main/java/org/apache/flink/runtime/jobmaster/slotpool/SlotPool.java

public class SlotPool extends RpcEndpoint implements SlotPoolGateway, AllocatedSlotActions {

	/** The interval (in milliseconds) in which the SlotPool writes its slot distribution on debug level. */
	private static final int STATUS_LOG_INTERVAL_MS = 60_000;

	private final JobID jobId;

	private final SchedulingStrategy schedulingStrategy;

	private final ProviderAndOwner providerAndOwner;

	/** All registered TaskManagers, slots will be accepted and used only if the resource is registered. */
	private final HashSet<ResourceID> registeredTaskManagers;

	/** The book-keeping of all allocated slots. */
	private final AllocatedSlots allocatedSlots;

	/** The book-keeping of all available slots. */
	private final AvailableSlots availableSlots;

	/** All pending requests waiting for slots. */
	private final DualKeyMap<SlotRequestId, AllocationID, PendingRequest> pendingRequests;

	/** The requests that are waiting for the resource manager to be connected. */
	private final HashMap<SlotRequestId, PendingRequest> waitingForResourceManager;

	/** Timeout for external request calls (e.g. to the ResourceManager or the TaskExecutor). */
	private final Time rpcTimeout;

	/** Timeout for releasing idle slots. */
	private final Time idleSlotTimeout;

	private final Clock clock;

	/** Managers for the different slot sharing groups. */
	protected final Map<SlotSharingGroupId, SlotSharingManager> slotSharingManagers;

	/** the fencing token of the job manager. */
	private JobMasterId jobMasterId;

	/** The gateway to communicate with resource manager. */
	private ResourceManagerGateway resourceManagerGateway;

	private String jobManagerAddress;

	//......

	/**
	 * Start the slot pool to accept RPC calls.
	 *
	 * @param jobMasterId The necessary leader id for running the job.
	 * @param newJobManagerAddress for the slot requests which are sent to the resource manager
	 */
	public void start(JobMasterId jobMasterId, String newJobManagerAddress) throws Exception {
		this.jobMasterId = checkNotNull(jobMasterId);
		this.jobManagerAddress = checkNotNull(newJobManagerAddress);

		// TODO - start should not throw an exception
		try {
			super.start();
		} catch (Exception e) {
			throw new RuntimeException("This should never happen", e);
		}

		scheduleRunAsync(this::checkIdleSlot, idleSlotTimeout);

		if (log.isDebugEnabled()) {
			scheduleRunAsync(this::scheduledLogStatus, STATUS_LOG_INTERVAL_MS, TimeUnit.MILLISECONDS);
		}
	}

	/**
	 * Check the available slots, release the slot that is idle for a long time.
	 */
	private void checkIdleSlot() {

		// The timestamp in SlotAndTimestamp is relative
		final long currentRelativeTimeMillis = clock.relativeTimeMillis();

		final List<AllocatedSlot> expiredSlots = new ArrayList<>(availableSlots.size());

		for (SlotAndTimestamp slotAndTimestamp : availableSlots.availableSlots.values()) {
			if (currentRelativeTimeMillis - slotAndTimestamp.timestamp > idleSlotTimeout.toMilliseconds()) {
				expiredSlots.add(slotAndTimestamp.slot);
			}
		}

		final FlinkException cause = new FlinkException("Releasing idle slot.");

		for (AllocatedSlot expiredSlot : expiredSlots) {
			final AllocationID allocationID = expiredSlot.getAllocationId();
			if (availableSlots.tryRemove(allocationID) != null) {

				log.info("Releasing idle slot [{}].", allocationID);
				final CompletableFuture<Acknowledge> freeSlotFuture = expiredSlot.getTaskManagerGateway().freeSlot(
					allocationID,
					cause,
					rpcTimeout);

				freeSlotFuture.whenCompleteAsync(
					(Acknowledge ignored, Throwable throwable) -> {
						if (throwable != null) {
							if (registeredTaskManagers.contains(expiredSlot.getTaskManagerId())) {
								log.debug("Releasing slot [{}] of registered TaskExecutor {} failed. " +
									"Trying to fulfill a different slot request.", allocationID, expiredSlot.getTaskManagerId(),
									throwable);
								tryFulfillSlotRequestOrMakeAvailable(expiredSlot);
							} else {
								log.debug("Releasing slot [{}] failed and owning TaskExecutor {} is no " +
									"longer registered. Discarding slot.", allocationID, expiredSlot.getTaskManagerId());
							}
						}
					},
					getMainThreadExecutor());
			}
		}

		scheduleRunAsync(this::checkIdleSlot, idleSlotTimeout);
	}

	//......
}
复制代码

• SlotPool在start方法里头，调用scheduleRunAsync方法，延时idleSlotTimeout调度执行checkIdleSlot；checkIdleSlot方法会挨个检查availableSlots的SlotAndTimestamp，判断当前时间与slotAndTimestamp.timestamp的时间差是否超过idleSlotTimeout，超过的话，则放入expiredSlots，之后对expiredSlots挨个进行availableSlots.tryRemove，然后调用TaskManagerGateway.freeSlot进行释放，之后再次调用scheduleRunAsync(this::checkIdleSlot, idleSlotTimeout)进行下一次的延时调度检测

RpcEndpoint

flink-release-1.7.2/flink-runtime/src/main/java/org/apache/flink/runtime/rpc/RpcEndpoint.java

public abstract class RpcEndpoint implements RpcGateway {
	//......

	/**
	 * Execute the runnable in the main thread of the underlying RPC endpoint, with
	 * a delay of the given number of milliseconds.
	 *
	 * @param runnable Runnable to be executed
	 * @param delay    The delay after which the runnable will be executed
	 */
	protected void scheduleRunAsync(Runnable runnable, Time delay) {
		scheduleRunAsync(runnable, delay.getSize(), delay.getUnit());
	}

	/**
	 * Execute the runnable in the main thread of the underlying RPC endpoint, with
	 * a delay of the given number of milliseconds.
	 *
	 * @param runnable Runnable to be executed
	 * @param delay    The delay after which the runnable will be executed
	 */
	protected void scheduleRunAsync(Runnable runnable, long delay, TimeUnit unit) {
		rpcServer.scheduleRunAsync(runnable, unit.toMillis(delay));
	}

	//......
}
复制代码

• RpcEndpoint提供了scheduleRunAsync，其最后调用的是rpcServer.scheduleRunAsync

小结

• slot.idle.timeout默认为HeartbeatManagerOptions.HEARTBEAT_TIMEOUT.defaultValue()，即50000L毫秒
• SlotPool在start方法里头，调用scheduleRunAsync方法，延时idleSlotTimeout调度执行checkIdleSlot；checkIdleSlot方法会挨个检查availableSlots的SlotAndTimestamp，判断当前时间与slotAndTimestamp.timestamp的时间差是否超过idleSlotTimeout，超过的话，则放入expiredSlots，之后对expiredSlots挨个进行availableSlots.tryRemove，然后调用TaskManagerGateway.freeSlot进行释放，之后再次调用scheduleRunAsync(this::checkIdleSlot, idleSlotTimeout)进行下一次的延时调度检测
• RpcEndpoint提供了scheduleRunAsync，其最后调用的是rpcServer.scheduleRunAsync

doc

• slot-idle-timeout