Status

...

...

JIRA:

...

Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).

Motivation

Flink jobs typically run in a distributed way. In large clusters, it’s very common for cluster nodes to encounter issues, such as insufficient disk space, bad hardware, network abnormalities. These problems may cause job failures. Flink will take care of the failures and redeploy the relevant tasks. However, due to data locality and limited resources, the new tasks are very likely to be redeployed to the same nodes, which will result in continuous task abnormalities and affect job progress. 

Currently, Flink users need to manually identify the problematic node and take it offline to solve this problem. But this approach has following disadvantages:

1. Taking a node offline can be a heavy process. Users may need to contact cluster administors to do this. The operation can even be dangerous and not allowed during some important business events.
2. Identifying and solving this kind of problems manually would be slow and a waste of human resources.

To solve this problem, we propose to introduce a blocklist mechanism for Flink to filter out problematic resources. Following two ways will be introduced to blocklist resources:

1. Manually specify the blocked resources through REST API. When users find abnormal nodes/TMs, they can manually blocklist them in this way.
2. Automatically detect abnormal resources and blocklist them. Users can specify a blocklist strategy which identifies abnormal resources. In the first version, we only introduce the blocklist strategy interface and a no-op implementation. In the future, we will introduce a configurable blocklist strategy and plugin mechanism to load user-defined blocklist strategy implementations.

Public Interfaces

We propose to introduce following configuration options for blocklist:

• cluster.resource-blocklist.enabled: Whether to enable blocklist mechanism.
• cluster.resource-blocklist.item.timeout: Timeout for blocked items in blocklist to be removed

Proposed Changes

In this design, two granularities of blocked resources are supported: task managers and nodes. A record of blocklist information is called a blocked item, which is generally generated by the scheduler according to the exception of the tasks. These blocked items will be recorded in a special component and affect the resource allocation of Flink clusters. However，the blocklist items are not permanent, there will be a timeout for it. Once an item times out, it will be removed, and the resource will become available again. The overall structure of the blocklist mechanism is shown in the figure below. 

Image Removed

In JM, there will be a component (JobMasterBlocklistHandler) responsible for generating blocked items according to the exceptions notified by Scheduler, and managing them. SlotPool will be aware of the blocklist information and filter out blocked resources when allocating slots.

In RM, there will also be a component (ResourceManagerBlocklistHandler) responsible for managing the cluster-level blocklist. SlotManager will be aware of the blocklist information  and filter out resources in the blocklist when allocating slots from registered task managers. Similarly, when the ResourceManagerDriver requests new resources from an external resource manager (Yarn or Kubernetes), it also needs to filter out the blocked nodes.

Blocked Item

A blocked item corresponds to the blocklist information of a specific task manager or node, including the following 5 fields:

1. type: The blocked item type, TASK_MANAGER or NODE
2. timestamp: The timestamp for creating this item, will be used to check timeout.
3. cause: The cause for creating this item.
4. identifier: The identifier of the blocked task manager or node.
5. action: The action when a task manager/node is marked as blocked, including:
   1. MARK_BLOCKED: Just mark the task manager or node as blocked. Future slots should not be allocated from the blocked task manager or node. But slots that are already allocated will not be affected.
   2. MARK_BLOCKED_AND_EVACUATE_TASKS: Mark the task manager or node as blocked, and evacuate all tasks on it. Evacuated tasks will be restarted on non-blocked task managers.

/**
 * This class represents a blocked item.
 *
 * @param <ID> Identifier of the blocked item.
 */
public abstract class BlockedItem<ID> {
    public BlockedItemType getType();

    public long getTimestamp();

    public BlocklistAction getAction();

    public Throwable getCause();

    public abstract ID getIdentifier();
}

/** This class represents a blocked node. */
public class BlockedNode extends BlockedItem<String> {
}

/** This class represents a blocked task manager. */
public class BlockedTaskManager extends BlockedItem<ResourceID> {
}

Blocklist on JobMaster

JobMasterBlocklistHandler

JobMasterBlocklistHandler is the component in JM responsible for generating and managing blocked items. It consists of three sub-components：BlocklistStrategy, BlocklistTracker and BlocklistContext. When receiving a new exception from Scheduler, the JobMasterBlocklistHandler will handle it as following:

1. Generate new blocked items by notifying the exception to the BlocklistStrategy.
2. Add the new blocked items to the BlocklistTracker. 
3. Synchronize the new blocked items to RM.
4. Perform blocklist actions on the resources via the BlocklistContext.

BlocklistStrategy

BlocklistStrategy is the component responsible for generating blocklist items according to the exceptions and their locations notified by Scheduler. We can introduce different BlocklistStrategy implementations to adapt to different scenarios. In the first version, we will introduce a no-op implementation as default implementation. In the future, we will introduce a configurable blocklist strategy and plugin mechanism to load user-defined blocklist strategy implementations, details will be described in Future improvements.

public interface BlocklistStrategy {
    /**
     * Generate blocked items according to the abnormal task's location and cause.
     *
     * @param locations the abnormal tasks’ locations.
     * @param cause the cause of blocked items.
     * @param timestamp the create timestamp of blocked items.
     * @return the generated blocked items.
     */
    Collection<BlockedItem<?>> generateBlockedItems(Collection<TaskManagerLocation> locations, Throwable cause, long timestamp);
}

BlocklistTracker

BlocklistTracker is the component responsible for tracking blocklist items. The tracker will regularly remove timeout blocked items.

public interface BlocklistTracker {
    /** Starts the blocklist tracker. */
    void start(ComponentMainThreadExecutor mainThreadExecutor);

    /**
     * Add new blocked items or update existing items.
     *
     * @param items The items to add or update
     * @return Newly added or updated items.
     */
    Collection<BlockedItem<?>> addNewBlockedItems(Collection<BlockedItem<?>> items);

    /** Returns whether the given task manager is blocked. */
    boolean isBlockedTaskManager(ResourceID resourceID);

    /** Get all blocked nodes. */
    Set<String> getBlockedNodes();

    /** Close the blocklist tracker. */
    void close();
}
     

BlocklistContext

BlocklistContext is the component responsible for performing blocklist actions on SlotPool, the details will be described in SlotPool.

public interface BlocklistContext {
    /** Perform the newly added or updated blocklist items on resources. */
    void blocklistResources(Collection<BlockedItem<?>> newlyAddedOrUpdatedItems);
}
   

public interface BlocklistHandler extends BlocklistTracker {
    /** Add a new blocked node. */
    void blockNode(String nodeId, BlacklistAction action, Throwable cause);

    /** Add a new blocked task manager. */
    void blockTaskManager(ResourceID taskManagerId, BlacklistAction action, Throwable cause);
 }

public interface JobMasterBlocklistHandler extends BlocklistHandler {
}

SlotPool

SlotPool should avoid allocating slots from blocked task managers. Blocked task managers include those directly blocked and those located on blocked nodes. To do that, our core idea is to keep the SlotPool in such a state: there is no slot in SlotPool that is free (no task assigned) and blocked. Details are as following:

1. When SlotPool starts, BlockedTaskManagerChecker will be passed in to check whether a task manager is blocked.
2. When receiving slot offers from blocked task managers (including task managers on blocked nodes), all offers should be rejected.
3. When a task manager is newly blocked, BlocklistContext will perform the following on SlotPool:
   1. If the action is MARK_BLOCKED, release all free(no task assigned) slots on the blocked task manager.
   2. If the action is MARK_BLOCKED_AND_EVACUATE_TASKS, release all slots on the blocked task manager.

4. When a slot state changes from reserved(task assigned) to free(no task assigned), it will check whether the corresponding task manager is blocked. If yes, release the slot.

public interface BlockedTaskManagerChecker {
    /**
     * Returns whether the given task manager is blocked.
     */
    boolean isBlockedTaskManager(ResourceID resourceID);
}


public interface SlotPoolService {

    /** Start the encapsulated slot pool implementation. */
    void start(
            JobMasterId jobMasterId,
            String address,
            ComponentMainThreadExecutor mainThreadExecutor,
            BlockedTaskManagerChecker blockedTaskManagerChecker)
            throws Exception;

   /**
     * Releases all slots belonging to the owning TaskExecutor if it has been registered.
     *
     * @param taskManagerId identifying the TaskExecutor
     * @param cause cause for failing the slots
     */
    void releaseSlotsOnTaskManager(ResourceID taskManagerId, Exception cause);

    /**
     * Releases all free slots belonging to the owning TaskExecutor if it has been registered.
     *
     * @param taskManagerId identifying the TaskExecutor
     * @param cause cause for failing the slots
     */
    void releaseFreeSlotsOnTaskManager(ResourceID taskManagerId, Exception cause);

    //...
}

Blocklist on ResourceManager

ResourceManagerBlocklistHandler

ResourceManagerBlocklistHandler is a new component introduced in RM for the blocklist mechanism. It has only one sub-component: BlocklistTracker, which is responsible for managing cluster-level blocked items.

public interface ResourceManagerBlocklistHandler extends BlocklistHandler {
}

SlotManager

SlotManager should filter out blocked resources when allocating registered resources. To do that, we need following changes:

1. When starting SlotManager, the BlockedTaskManagerChecker will be passed in to check whether a registered task manager is blocked.
2. When trying to fulfill the slot requirements by registered task managers, the blocked ones will be filtered out.
3. SlotManager will request new task managers from external resource managers if the registered resources cannot fulfill the requirements. The blocklist also takes effect when requesting new task managers, the details will be described in ResourceManagerDriver.

In order to support speculative execution for batch jobs(FLIP-168), we need a mechanism to block resources on nodes where the slow tasks are located. We propose to introduce a blocklist mechanism as follows:  Once a node is marked as blocked, future slots should not be allocated from the blocked node, but the slots that are already allocated will not be affected.

Proposed Changes

We introduce a new data structure: BlockedNode to record the information of a blocked node. This information will be recorded in a special component and affect the resource allocation of Flink clusters. However，the blocking of nodes is not permanent, there will be a timeout for it. Once it times out, the nodes will become available again. The overall structure of the blocklist mechanism is shown in the figure below. 

Image Added

In JM, there will be a component (JobMasterBlocklistHandler) responsible for managing all BlockedNode(s) and performing them on SlotPool. 

In RM, there will also be a component (ResourceManagerBlocklistHandler) responsible for managing the cluster-level blocklist. SlotManager will be aware of the blocklist information and filter out resources in the blocklist when allocating slots from registered task managers. Similarly, when the ResourceManagerDriver requests new resources from an external resource manager (Yarn or Kubernetes), it also needs to filter out the blocked nodes.

BlockedNode

A BlockedNode corresponds to the blocked information of a specific node, including the following 4 fields:

1. id: The identifier of the node.
2. startTimestamp: The start time of the blocking.
3. endTimestamp: The end time of the blocking (at which time the node will become available again). 
4. cause: The cause for blocking this node.

/**
 * This class represents a blocked node.
 */
public class BlockedNode {
    
    public long getStartTimestamp();

    public long getEndTimestamp();
    
    public String getCause();

    public String getId();
}

Blocklist on JobMaster

JobMasterBlocklistHandler

JobMasterBlocklistHandler is the component in JM responsible for managing all blocked node information and performing them on SlotPool. It consists of two sub-components：BlocklistTracker and BlocklistContext. When receiving a new blocked node information, the JobMasterBlocklistHandler will handle it as following:

1. Add the new blocked nodes to the BlocklistTracker. 
2. Synchronize the new blocked nodes to RM.
3. Block the resources on newly added nodes via the BlocklistContext.

BlocklistTracker

BlocklistTracker is the component responsible for tracking blocked nodes. The tracker will regularly remove timeout blocked nodes.

public interface BlocklistTracker {

    /**
     * Add or update blocked nodes.
     *
     * @param nodes The nodes to add or update
     * @return Newly added or updated nodes.
     */
    Collection<BlockedNode> addNewBlockedNodes(Collection<BlockedNode> nodes);

    /** Returns whether the given task manager is located on blocked nodes. */
    boolean isBlockedTaskManager(ResourceID taskManagerId);

public interface SlotManager {

    /** StartsGet theall slot manager with the given leader id and resource manager actionsblocked nodes. */
    voidSet<String> startgetBlockedNodes();

    /** Remove the timeout     ResourceManagerId newResourceManagerId,nodes. */
    void removeTimeoutNodes();
}
     

BlocklistContext

BlocklistContext is the component responsible for blocking slots in SlotPool, the details will be described in SlotPool.

public interface BlocklistContext  Executor newMainThreadExecutor,{
    /** Block resources on the newly  added ResourceActions newResourceActions,nodes. */
    void blockResources(Collection<BlockedNode> newlyAddedOrUpdatedNodes);
}
   

public interface BlocklistHandler BlockedTaskManagerChecker newBlockedTaskManagerChecker);{

    //...
}

ResourceManagerDriver

When deploying Flink on Yarn or Kubernetes, the ResourceManagerDriver is responsible for requesting new task managers from the external resource manager. To avoid allocating task managers from blocked nodes, we need to pass the information of blocked nodes to ResourceManagerDriver at its initialization. 

ResourceManagerDriver uses following APIs to tell external resource managers about the information of blocklist nodes:

1. Yarn: AMRMClient#updateBlacklist 
2. Kubernetes: NodeAffinity

public interface BlockedNodeRetriever {

    /** Retrieve** Add a new blocked node. */
    void blockNode(String nodeId, String cause, long startTimestamp, long endTimestamp);

    /** Returns whether the given task manager is located on blocked nodes. */
    Set<String>boolean getBlockedNodesisBlockedTaskManager(ResourceID taskManagerId);
}

public interface ResourceManagerDriver {

    /** InitializeGet theall deploymentblocked specific componentsnodes. */
    voidSet<String> initializegetBlockedNodes(
);            ResourceEventHandler<WorkerType> resourceEventHandler,
            ScheduledExecutor mainThreadExecutor,
            Executor ioExecutor,
            BlockedNodeRetriever blockedNodeRetriever)
            throws Exception;

    //...
}

Synchronize Blocklist between JM & RM

The newly added or updated blocked items will be synchronized between JM and RM via RPC: 

1. Once a few blocked items are newly added (or updated) to the JobMasterBlocklistHandler, RM will be notified of these items via ResourceManagerGateway#notifyNewBlockedItems.
2. When RM receives the blocked items notified by a JM, it will add them into ResourceManagerBlocklistHandler, and notify all JMs of the successfully added (or updated) items through JobMasterGateway#notifyNewBlockedItems. 
3. Similarly, when JM receives the blocked items notified by RM, it will also add them to JobMasterBlocklistHandler.

 
}

public interface JobMasterBlocklistHandler extends BlocklistHandler {
}

SlotPool

SlotPool should avoid allocating slots that located on blocked nodes. To do that, our core idea is to keep the SlotPool in such a state: there is no slot in SlotPool that is free (no task assigned) and located on blocked nodes. Details are as following:

1. When receiving slot offers from task managers located on blocked nodes, all offers should be rejected.
2. When a node is newly blocked, we should release all free(no task assigned) slots on it. We need to find all task managers on blocked nodes and release all free slots on them by SlotPoolService#releaseFreeSlotsOnTaskManager.

3. When a slot state changes from reserved(task assigned) to free(no task assigned), it will check whether the corresponding task manager is blocked. If yes, release the slot.

We will introduce a new slot pool implementation: BlocklistSlotPool, which extends the DefaultDeclarativeSlotPool and overrides some methods to implement the blocklist-related functions described above, and the blocklist information will be passed in the constructor. This new implementation will be only used when the blocklist is enabled.

public interface SlotPoolService {
    /**
     * Releases all free slots belonging to the owning TaskExecutor if it has been registered.
     *
     * @param taskManagerId identifying the TaskExecutor
     * @param cause cause for failing the slots
    

public interface BlocklistListener {

    /** Notify new blocked items. */
    void notifyNewBlockedItems(Collection<BlockedItem<?>> newItemsreleaseFreeSlotsOnTaskManager(ResourceID taskManagerId, Exception cause);
}

public interface JobManagerGateway
    //...
}

public class BlocklistSlotPool extends BlocklistListenerDefaultDeclarativeSlotPool {
    // ...
}

public interface ResourceManagerGateway extends BlocklistListener {     
    //...
}  

Enrich TaskManagerLocation with node information

...

Blocklist on ResourceManager

ResourceManagerBlocklistHandler

ResourceManagerBlocklistHandler is a new component introduced in RM for the blocklist mechanism. It has only one sub-component: BlocklistTracker, which is responsible for managing cluster-level blocklist.

public interface ResourceManagerBlocklistHandler extends BlocklistHandler {
}

SlotManager

SlotManager should filter out blocked resources when allocating registered resources. To do that, we need following changes:

1. When starting SlotManager, the BlockedTaskManagerChecker will be passed in to check whether a registered task manager is located on blocked nodes.
2. When trying to fulfill the slot requirements by registered task managers, the task managers located on blocked nodes will be filtered out.
3. SlotManager will request new task managers from external resource managers if the registered resources cannot fulfill the requirements. The blocklist also takes effect when requesting new task managers, the details will be described in ResourceManagerDriver.

/** This checker helps to query whether a given task manager is located on blocked nodes. */
public interface BlockedTaskManagerChecker {public class TaskManagerLocation {

    public String getNodeId();

    //...
}

Metrics

We propose to introduce following Gauge metrics to ResourceManagerMetricGroup:

1. numBlockedTaskMangers: The number of currently blocked task managers (including task managers on blocked nodes)
2. numBlockedNodes: The number of currently blocked nodes

REST API

We will introduce following REST APIs for blocklist mechanism:

1. REST API for querying blocklist information.
2. REST API for adding new blocked items.
3. REST API for removing existing blocked items.

Query

Add a REST API to obtain blocklist information. Each request will return all current blocklist items, which are obtained from ResourceManagerBlocklistHandler.

GET: http://{jm_rest_address:port}/blocklist

Request: {}

Response:

** Returns whether the given task manager is located on blocked nodes. */
    boolean isBlockedTaskManager(ResourceID taskManagerId);
}

public interface SlotManager {

    /** Starts the slot manager with the given leader id and resource manager actions. */
    void start(
            ResourceManagerId newResourceManagerId,
            Executor newMainThreadExecutor,
            ResourceActions newResourceActions,
            BlockedTaskManagerChecker newBlockedTaskManagerChecker);

    //...
}

ResourceManagerDriver

When deploying Flink on Yarn or Kubernetes, the ResourceManagerDriver is responsible for requesting new task managers from the external resource manager. To avoid allocating task managers from blocked nodes, we need to pass the information of blocked nodes to ResourceManagerDriver at its initialization. 

ResourceManagerDriver uses following APIs to tell external resource managers about the information of blocked nodes:

1. Yarn: AMRMClient#updateBlacklist 
2. Kubernetes: NodeAffinity

public interface BlockedNodeRetriever {

    /** Retrieve blocked nodes. */
    Set<String> getBlockedNodes();
}

public interface ResourceManagerDriver {

    /** Initialize the deployment specific components. */
    void initialize(
 

{
  /** This group only contains directly blocked task managers */
  "blockedTaskManagers": [
      {
          "id" : "container_XXX_000002",
          "timestamp" : "XXX",
          "action" : "MARK_BLOCKED"
      },
      {
          "id" : "container_XXX_000003",
          "timestamp" : "XXX",
          "action" : "MARK_BLOCKED"
      }, 
      ...
  ],
  "blockedNodes": [
      {
          "id" :ResourceEventHandler<WorkerType> "node1"resourceEventHandler,
          "timestamp"  :ScheduledExecutor "XXX"mainThreadExecutor,
           "action" : "MARK_BLOCKED"Executor ioExecutor,
          "taskManagers" : [“container_XXX_000004”, “container_XXX_000005”, …]
 BlockedNodeRetriever blockedNodeRetriever)
       },
      ...
  ]
}

In order to get blocklist information, we need to add an interface to ResourceManagerGateway:

public interface ResourceManagerGateway {
   CompletableFuture<BlocklistInfo> requestBlocklist(@RpcTimeout Time timeout);
   // ...
}

Add

POST: http://{jm_rest_address:port}/blocklist/nodes

POST: http://{jm_rest_address:port}/blocklist/taskmanagers

Request:

{
  [
      {
          "id" : "nodeXXX/container_XXX",
          "action" : "MARK_BLOCKED"
      },
      {           
          "id" : "nodeXXX/container_XXX",
          "action" : "MARK_BLOCKED"
      }, 
      ...
  ]
}

Response: {}

Remove

DELETE: http://{jm_rest_address:port}/blocklist/node/<id>/<action>

DELETE: http://{jm_rest_address:port}/blocklist/taskmanager/<id>/<action>

Request: {}

Response: {}

throws Exception;

    //...
}

Synchronize Blocklist between JM & RM

The newly added/updated blocked nodes will be synchronized between JM and RM via RPC: 

1. Once a few blocked nodes are newly added/updated to the JobMasterBlocklistHandler, RM will be notified of these nodes via ResourceManagerGateway#notifyNewBlockedNodes.
2. When RM receives the blocked nodes notified by a JM, it will add them into ResourceManagerBlocklistHandler, and notify all JMs of the successfully added/updated nodes through JobMasterGateway#notifyNewBlockedNodes. 
3. Similarly, when JM receives the blocked nodes notified by RM, it will also add them to JobMasterBlocklistHandler.

public interface BlocklistListener {

    /** Notify newly added/updated blocked nodes. */
    void notifyNewBlockedNodes(Collection<BlockedNode> newlyAddedOrUpdatedNodes);
}

public interface JobManagerGateway extends BlocklistListener {
    //...
}

public interface ResourceManagerGateway extends BlocklistListener {     
    //...
}  

Enrich TaskManagerLocation with node information

In order to support blocked nodes, it is necessary to know the node where the task is located. To do that, we need to add a node identifier into TaskManagerLocation. This node identifier should be set by the resource manager when requesting new task managers.

public class TaskManagerLocation {

    public String getNodeId();

    //...
}

Compatibility, Deprecation, and Migration Plan

The blocklist mechanism will be an optional feature which the user has to activate explicitly by setting the config option cluster.resource-blocklist.enabled: true. This entails is only used when speculative execution is enabled, which entails that Flink's default behavior won't change.

Future improvements

Introduce a configurable blocklist strategy implementation

We intend to introduce a configurable blocklist strategy in the future. Users can specify  exceptions for which the strategy should blocklist the resources. This requires adding several configuration options to configure exceptions and the generated blocklist item, which requires further design and discussion.

Introduce plugin mechanism for blocklist strategy

Limitations

No integration with Flink's web UI

This FLIP does not modify the web UI, but it's in our plan. Currently, we have a preliminary idea of the UI intergration, including improve the slots information displaying and add a page to show the blocklist information.

No support for JM failover

Currently, the blocklist information will be lost after JM failover.  In the future, we may persist the blocklist information in HA to support JM failoverWe may introduce a plugin mechanism in the future, which allows users to load their own blocklist strategy implementations. This means that BlocklistStrategy needs to be opened to users as a public interface, which requires more thought and discussion.

Test Plan

1. The changes will be covered by unit and IT cases.
2. Test the functionality in a real Standanlone/Yarn/Kubernetes cluster.

...