Authors: The Yunikorn Scheduler Authors
To define a standard interface that can be used by different types of resource management systems such as YARN/K8s.
Possible use cases:
Highlights:
Interface and messages generic definition.
The syntax used for the declarations is proto3
. The definition currently only provides go related info.
syntax = "proto3"; package si.v1; import "google/protobuf/descriptor.proto"; option go_package = "lib/go/si"; extend google.protobuf.FieldOptions { // Indicates that a field MAY contain information that is sensitive // and MUST be treated as such (e.g. not logged). bool si_secret = 1059; }
There are two kinds of interfaces, the first one is RPC based communication, the second one is API based.
RPC based, the gRPC framework is used, will be useful when scheduler has to be deployed as a remote process. For example when we need to deploy scheduler support multiple remote clusters. A second example is when there is a cross language integration, like between Java and Go.
Unless specifically required we strongly recommend the use of the API based interface to avoid the overhead of the RPC serialization and de-serialization.
There are three sets of RPCs:
Currently only the design and implementation for the Scheduler Service is provided.
service Scheduler { // Register a RM, if it is a reconnect from previous RM the call will // trigger a cleanup of all in-memory data and resync with RM. rpc RegisterResourceManager (RegisterResourceManagerRequest) returns (RegisterResourceManagerResponse) { } // Update Scheduler status (this includes node status update, allocation request // updates, etc. And receive updates from scheduler for allocation changes, // any required status changes, etc. // Update allocation request rpc UpdateAllocation(stream AllocationRequest) returns (stream AllocationResponse) { } // Update application request rpc UpdateApplication(stream ApplicationRequest) returns (stream ApplicationResponse) { } // Update node info rpc UpdateNode(stream NodeRequest) returns (stream NodeResponse) { } }
The reason of using bi-directional streaming gRPC is, according to performance benchmark: https://grpc.io/docs/guides/benchmarking.html latency is close to 0.5 ms. The same performance benchmark shows streaming QPS can be 4x of non-streaming RPC. Considering scheduler needs both throughput and better latency, we go with streaming API for scheduler related decisions.
The API interface only relies on the message definition and not on other generated code as the RPC Interface does. Below is an example of the Scheduler Service as defined in the RPC. The SchedulerAPI is bi-directional and can be a-synchronous. For the asynchronous cases the API requires a callback interface to be implemented in the resource manager. The callback must be provided to the scheduler as part of the registration.
package api import "github.com/apache/yunikorn-scheduler-interface/lib/go/si" type SchedulerAPI interface { // Register a new RM, if it is a reconnect from previous RM, cleanup // all in-memory data and resync with RM. RegisterResourceManager(request *si.RegisterResourceManagerRequest, callback ResourceManagerCallback) (*si.RegisterResourceManagerResponse, error) // Update allocation request UpdateAllocation(request *si.AllocationRequest) error // Update application request UpdateApplication(request *si.ApplicationRequest) error // Update node info UpdateNode(request *si.NodeRequest) error // Notify scheduler to reload configuration and hot-refresh in-memory state based on configuration changes UpdateConfiguration(request *si.UpdateConfigurationRequest) error } // RM side needs to implement this API type ResourceManagerCallback interface { //Receive Allocation Update Response UpdateAllocation(response *si.AllocationResponse) error //Receive Application Update Response UpdateApplication(response *si.ApplicationResponse) error //Receive Node Update Response UpdateNode(response *si.NodeResponse) error // Run a certain set of predicate functions to determine if a proposed allocation // can be allocated onto a node. Predicates(args *si.PredicatesArgs) error // Run predicate functions to determine if a proposed allocation can be allocated // onto a node after preemption. The request contains a list of allocations to // speculatively remove. PreemptionPredicates(args *si.PreemptionPredicatesArgs) *si.PreemptionPredicatesResponse // This plugin is responsible for transmitting events to the shim side. // Events can be further exposed from the shim. SendEvent(events []*si.EventRecord) // Scheduler core can update container scheduling state to the RM, // the shim side can determine what to do incorporate with the scheduling state // update container scheduling state to the shim side // this might be called even the container scheduling state is unchanged // the shim side cannot assume to only receive updates on state changes // the shim side implementation must be thread safe UpdateContainerSchedulingState(request *si.UpdateContainerSchedulingStateRequest) } // RM can additionally implement this API to provide information during state dumps type StateDumpPlugin interface { // This plugin is responsible for returning a JSON representation of the state of the shim GetStateDump() (string, error) }
Lifecycle of RM-Scheduler communication
Status of RM in scheduler: Connection timeout +-------+ +-------+ loss +-------+ +---------+ |init |+---->|Running|+---->|Paused |+---->| Stopped | +-------+ +----+--+ +-------+ +---------+ RM register | ^ with scheduler | | +-----------------------------+ RM voluntarilly Shutdown
When a new RM starts, fails, it will register with scheduler. In some cases, scheduler can ask RM to re-register because of connection issues or other internal issues.
message RegisterResourceManagerRequest { // An ID which can uniquely identify a RM **cluster**. (For example, if a RM cluster has multiple manager instances for HA purpose, they should use the same information when do registration). // If RM register with the same ID, all previous scheduling state in memory will be cleaned up, and expect RM report full scheduling state after registration. string rmID = 1; // Version of RM scheduler interface client. string version = 2; // Policy group name: // This defines which policy to use. Policy should be statically configured. (Think about network security group concept of ec2). // Different RMs can refer to the same policyGroup if their static configuration is identical. string policyGroup = 3; // Pass the build information of k8shim to core. map<string, string> buildInfo = 4; // Pass the serialized configuration for this policyGroup to core. string config = 5; // Additional configuration key/value pairs for configuration not related to the policyGroup. map<string, string> extraConfig = 6; } // Upon success, scheduler returns RegisterResourceManagerResponse to RM, otherwise RM receives exception. message RegisterResourceManagerResponse { // Intentionally empty. }
Below is overview of how scheduler/RM keep connection and updates.
message AllocationRequest { // New allocation requests or replace existing allocation request (if allocationID is same) repeated AllocationAsk asks = 1; // Allocations can be released. AllocationReleasesRequest releases = 2; // ID of RM, this will be used to identify which RM of the request comes from. string rmID = 3; // Existing allocations to be added. repeated Allocation allocations = 4; } message ApplicationRequest { // RM should explicitly add application when allocation request also explictly belongs to application. // This is optional if allocation request doesn't belong to a application. (Independent allocation) repeated AddApplicationRequest new = 1; // RM can also remove applications, all allocation/allocation requests associated with the application will be removed repeated RemoveApplicationRequest remove = 2; // ID of RM, this will be used to identify which RM of the request comes from. string rmID = 3; } message NodeRequest { // New node can be scheduled. If a node is notified to be "unscheduable", it needs to be part of this field as well. repeated NodeInfo nodes = 1; // ID of RM, this will be used to identify which RM of the request comes from. string rmID = 2; } message AllocationResponse { // New allocations repeated Allocation new = 1; // Released allocations, this could be either ack from scheduler when RM asks to terminate some allocations. // Or it could be decision made by scheduler (such as preemption or timeout). repeated AllocationRelease released = 2; // Released allocation asks(placeholder), when the placeholder allocation times out repeated AllocationAskRelease releasedAsks = 3; // Rejected allocation requests repeated RejectedAllocationAsk rejected = 4; // Rejected allocations repeated RejectedAllocation rejectedAllocations = 5; } message ApplicationResponse { // Rejected Applications repeated RejectedApplication rejected = 1; // Accepted Applications repeated AcceptedApplication accepted = 2; // Updated Applications repeated UpdatedApplication updated = 3; } message NodeResponse { // Rejected Node Registrations repeated RejectedNode rejected = 1; // Accepted Node Registrations repeated AcceptedNode accepted = 2; } message UpdatedApplication { // The application ID that was updated string applicationID = 1; // State of the application string state = 2; // Timestamp of the state transition int64 stateTransitionTimestamp = 3; // Detailed message string message = 4; } message RejectedApplication { // The application ID that was rejected string applicationID = 1; // A human-readable reason message string reason = 2; } message AcceptedApplication { // The application ID that was accepted string applicationID = 1; } message RejectedNode { // The node ID that was rejected string nodeID = 1; // A human-readable reason message string reason = 2; } message AcceptedNode { // The node ID that was accepted string nodeID = 1; }
Lifecycle of AllocationAsk:
Rejected by Scheduler +-------------------------------------------+ | | | v +-------+---+ Asked +-----------+Scheduler or,+-----------+ |Initial +------->|Pending |+----+----+->|Rejected | +-----------+By RM +-+---------+ Asked by RM +-----------+ + | v +-----------+ |Allocated | +-----------+
Lifecycle of Allocations:
+--Allocated by v Scheduler +-----------+ +------------+ |Allocated |+------ |Completed | +---+-------+ Stoppe +------------+ | by RM | +------------+ +--------------->|Preempted | + Preempted by +------------+ | Scheduler | | | +------------+ +--------------->|Expired | Timeout +------------+ (Part of Allocation ask)
Common fields for allocation:
// A sparse map of resource to Quantity. message Resource { map<string, Quantity> resources = 1; } // Quantity includes a single int64 value message Quantity { int64 value = 1; }
Allocation ask:
message AllocationAsk { // Allocation key is used by both of scheduler and RM to track allocations. // It doesn't have to be same as RM's internal allocation id (such as Pod name of K8s or ContainerID of YARN). // Allocations from the same AllocationAsk which are returned to the RM at the same time will have the same allocationKey. // The request is considered an update of the existing AllocationAsk if an ALlocationAsk with the same allocationKey // already exists. string allocationKey = 1; // The application ID this allocation ask belongs to string applicationID = 2; // The partition the application belongs to string partitionName = 3; // The amount of resources per ask Resource resourceAsk = 4; // Maximum number of allocations int32 maxAllocations = 5; // Priority of ask int32 priority = 6; // Execution timeout: How long this allocation will be terminated (by scheduler) // once allocated by scheduler, 0 or negative value means never expire. int64 executionTimeoutMilliSeconds = 7; // A set of tags for this spscific AllocationAsk. Allocation level tags are used in placing this specific // ask on nodes in the cluster. These tags are used in the PlacementConstraints. // These tags are optional. map<string, string> tags = 8; // The name of the TaskGroup this ask belongs to string taskGroupName = 9; // Is this a placeholder ask (true) or a real ask (false), defaults to false // ignored if the taskGroupName is not set bool placeholder = 10; // Is this ask the originator of the application? bool Originator = 11; // The preemption policy for this ask PreemptionPolicy preemptionPolicy = 12; }
Preemption policy:
message PreemptionPolicy { // Opt-out from preemption bool allowPreemptSelf = 1; // Allow preemption of other tasks with same or lower priority bool allowPreemptOther = 2; }
Application requests:
message AddApplicationRequest { // The ID of the application, must be unique string applicationID = 1; // The queue this application is requesting. The scheduler will place the application into a // queue according to policy, taking into account the requested queue as per the policy. string queueName = 2; // The partition the application belongs to string partitionName = 3; // The user group information of the application owner UserGroupInformation ugi = 4; // A set of tags for the application. These tags provide application level generic inforamtion. // The tags are optional and are used in placing an appliction or scheduling. // Application tags are not considered when processing AllocationAsks. map<string, string> tags = 5; // Execution timeout: How long this application can be in a running state // 0 or negative value means never expire. int64 executionTimeoutMilliSeconds = 6; // The total amount of resources gang placeholders will request Resource placeholderAsk = 7; // Gang scheduling style can be hard (the application will fail after placeholder timeout) // or soft (after the timeout the application will be scheduled as a normal application) string gangSchedulingStyle = 8; } message RemoveApplicationRequest { // The ID of the application to remove string applicationID = 1; // The partition the application belongs to string partitionName = 2; }
User information: The user that owns the application. Group information can be empty. If the group information is empty the groups will be resolved by the scheduler when needed.
message UserGroupInformation { // the user name string user = 1; // the list of groups of the user, can be empty repeated string groups = 2; }
The Allocation message is used in two cases:
message Allocation { // AllocationKey from AllocationAsk string allocationKey = 1; // Allocation tags from AllocationAsk map<string, string> allocationTags = 2; // UUID of the allocation string UUID = 3; // Resource for each allocation Resource resourcePerAlloc = 5; // Priority of ask int32 priority = 6; // Node which the allocation belongs to string nodeID = 8; // The ID of the application string applicationID = 9; // Partition of the allocation string partitionName = 10; // The name of the TaskGroup this allocation belongs to string taskGroupName = 11; // Is this a placeholder allocation (true) or a real allocation (false), defaults to false // ignored if the taskGroupName is not set bool placeholder = 12; reserved 7; reserved "queueName"; }
message AllocationReleasesRequest { // The allocations to release repeated AllocationRelease allocationsToRelease = 1; // The asks to release repeated AllocationAskRelease allocationAsksToRelease = 2; } enum TerminationType { UNKNOWN_TERMINATION_TYPE = 0;//TerminationType not set STOPPED_BY_RM = 1; // Stopped or killed by ResourceManager (created by RM) TIMEOUT = 2; // Timed out based on the executionTimeoutMilliSeconds (created by core) PREEMPTED_BY_SCHEDULER = 3; // Preempted allocation by scheduler (created by core) PLACEHOLDER_REPLACED = 4; // Placeholder allocation replaced by real allocation (created by core) } // Release allocation: this is a bidirectional message. The Terminationtype defines the origin, or creator, // as per the comment. The confirmation or response from the receiver is the same message with the same // termination type set. message AllocationRelease { // The name of the partition the allocation belongs to string partitionName = 1; // The application the allocation belongs to string applicationID = 2; // The UUID of the allocation to release, if not set all allocations are released for // the applicationID string UUID = 3; // Termination type of the released allocation TerminationType terminationType = 4; // human-readable message string message = 5; // AllocationKey from AllocationAsk string allocationKey = 6; } // Release ask message AllocationAskRelease { // Which partition to release the ask from, required. string partitionName = 1; // optional, when this is set, filter allocation key by application id. // when application id is set and allocationKey is not set, release all allocations key under the application id. string applicationID = 2; // optional, when this is set, only release allocation ask by specified string allocationKey = 3; // Termination type of the released allocation ask TerminationType terminationType = 4; // For human-readable message string message = 5; }
State transition of node:
+-----------+ +--------+ +-------+ |SCHEDULABLE|+-------->|DRAINING|+---------->|REMOVED| +-----------+ +--------+ +-------+ ^ Asked by + Aasked by | RM to DRAIN | RM to REMOVE | | +---------------------+ Asked by RM to SCHEDULE again
See protocol below:
During new node registration with the scheduler, request will be rejected if the node exist already. While updating registered node with the scheduler, request will fail if the node doesn't exist.
message NodeInfo { // Action from RM enum ActionFromRM { //ActionFromRM not set UNKNOWN_ACTION_FROM_RM = 0; // Create Node as initially schedulable. CREATE = 1; // Update node resources, attributes. UPDATE = 2; // Do not allocate new allocations on the node. DRAIN_NODE = 3; // Decomission node, it will immediately stop allocations on the node and // remove the node from schedulable lists. DECOMISSION = 4; // From Draining state to SCHEDULABLE state. // If node is not in draining state, error will be thrown DRAIN_TO_SCHEDULABLE = 5; // Create Node as initially draining (i.e. unschedulable). Before scheduling can proceed, // DRAIN_TO_SCHEDULABLE must be called. CREATE_DRAIN = 6; } // ID of node, the node must exist to be updated string nodeID = 1; // Action to perform by the scheduler ActionFromRM action = 2; // New attributes of node, which will replace previously reported attribute. map<string, string> attributes = 3; // new schedulable resource, scheduler may preempt allocations on the // node or schedule more allocations accordingly. Resource schedulableResource = 4; // when the scheduler is co-exist with some other schedulers, some node // resources might be occupied (allocated) by other schedulers. Resource occupiedResource = 5; // Allocated resources, this will be added when node registered to RM (recovery) repeated Allocation existingAllocations = 6; }
The following is the feedback produced from the scheduler to the RM:
Rejected allocation ask:
message RejectedAllocationAsk { // the ID of the allocation ask string allocationKey = 1; // The ID of the application string applicationID = 2; // A human-readable reason message string reason = 3; }
Rejected allocation:
message RejectedAllocation { // the ID of the allocation string allocationKey = 1; // The ID of the application string applicationID = 2; // A human-readable reason message string reason = 3; }
Scheduler Interface attributes start with the si prefix. Such constants are for example known attribute names for nodes and applications.
// Constants for node attributes const ( ARCH = "si/arch" HostName = "si/hostname" RackName = "si/rackname" OS = "si/os" InstanceType = "si/instance-type" FailureDomainZone = "si/zone" FailureDomainRegion = "si/region" LocalImages = "si/local-images" NodePartition = "si/node-partition" ) // Constants for allocation attributes const ( ApplicationID = "si/application-id" ContainerImage = "si/container-image" ContainerPorts = "si/container-ports" )
Allocation tags are key-value pairs, where the key should contain a domain, and optionally a group part. These parts should precede the name of the key (and should be in that order) and separated by a “/” character. Example allocation key: kubernetes.io/meta/namespace
.
// Constants for allocation tags const ( // Domains DomainK8s = "kubernetes.io/" DomainYuniKorn = "yunikorn.apache.org/" // Groups GroupMeta = "meta/" GroupLabel = "label/" GroupAnnotation = "annotation/" // Keys KeyPodName = "podName" KeyNamespace = "namespace" KeyRequiredNode = "requiredNode" // Pods CreationTime = "creationTime" )
Miscellaneous constants for resources and other values.
// Constants for Core and Shim const ( Memory = "memory" CPU = "vcore" AppTagNamespaceResourceQuota = "namespace.resourcequota" AppTagNamespaceResourceGuaranteed = "namespace.resourceguaranteed" AppTagStateAwareDisable = "application.stateaware.disable" AppTagCreateForce = "application.create.force" NodeReadyAttribute = "ready" )
SchedulerPlugin is a way to extend scheduler capabilities. Scheduler shim can implement such plugin and register itself to yunikorn-core, so plugged function can be invoked in the scheduler core.
message PredicatesArgs { // allocation key identifies a container, the predicates function is going to check // if this container is eligible to be placed ont to a node. string allocationKey = 1; // the node ID the container is assigned to. string nodeID = 2; // run the predicates for alloactions (true) or reservations (false) bool allocate = 3; } message PreemptionPredicatesArgs { // the allocation key of the container to check string allocationKey = 1; // the node ID the container should be attempted to be scheduled on string nodeID = 2; // a list of existing allocations that should be tentatively removed before checking repeated string preemptAllocationKeys = 3; // index of last allocation in starting attempt (first attempt should be 0..startIndex) int32 startIndex = 4; } message PreemptionPredicatesResponse { // whether or not container will schedule on the node bool success = 1; // index of last allocation which was removed before success (ignored during failure) int32 index = 2; } message UpdateContainerSchedulingStateRequest { // container scheduling states enum SchedulingState { //SchedulingState not set UNKNOWN_SCHEDULING_STATE = 0; // the container is being skipped by the scheduler SKIPPED = 1; // the container is scheduled and it has been assigned to a node SCHEDULED = 2; // the container is reserved on some node, but not yet assigned RESERVED = 3; // scheduler has visited all candidate nodes for this container // but non of them could satisfy this container's requirement FAILED = 4; } // application ID string applicationID = 1; // allocation key used to identify a container. string allocationKey = 2; // container scheduling state SchedulingState state = 3; // an optional plain message to explain why it is in such state string reason = 4; } message UpdateConfigurationRequest { // RM ID to update string rmID = 2; // PolicyGroup to update string policyGroup = 3; // New configuration to update string config = 4; // Additional configuration key/value pairs for configuration not related to the policyGroup. map<string, string> extraConfig = 5; reserved 1; reserved "configs"; }
The Event Cache is a SchedulerPlugin that exposes events about scheduler objects aiming to help the end user to see these events from the shim side. An event is sent to the shim through the callback as an EventRecord
. An EventRecord
consists of the following fields:
message EventRecord { enum Type { //EventRecord Type not set UNKNOWN_EVENTRECORD_TYPE = 0; REQUEST = 1; APP = 2; NODE = 3; QUEUE = 4; } enum ChangeType { NONE = 0; SET = 1; ADD = 2; REMOVE = 3; } enum ChangeDetail { DETAILS_NONE = 0; REQUEST_CANCEL = 100; // Request cancelled by the RM REQUEST_ALLOC = 101; // Request allocated REQUEST_TIMEOUT = 102; // Request cancelled due to timeout APP_ALLOC = 200; // Allocation changed APP_REQUEST = 201; // Request changed APP_REJECT = 202; // Application rejected on create APP_NEW = 203; // Application added with state new APP_ACCEPTED = 204; // State change to accepted APP_STARTING = 205; // State change to starting APP_RUNNING = 206; // State change to running APP_COMPLETING = 207; // State change to completing APP_COMPLETED = 208; // State change to completed APP_FAILING = 209; // State change to failing APP_FAILED = 210; // State change to failed APP_RESUMING = 211; // State change to resuming APP_EXPIRED = 212; // State change to expired NODE_DECOMISSION = 300; // Node removal NODE_READY = 301; // Node ready state change NODE_SCHEDULABLE = 302; // Node schedulable state change (cordon) NODE_ALLOC = 303; // Allocation changed NODE_CAPACITY = 304; // Capacity changed NODE_OCCUPIED = 305; // Occupied resource changed NODE_RESERVATION = 306; // Reservation/unreservation occurred QUEUE_CONFIG = 400; // Managed queue update or removal QUEUE_DYNAMIC = 401; // Dynamic queue update or removal QUEUE_TYPE = 402; // Queue type change QUEUE_MAX = 403; // Max resource changed QUEUE_GUARANTEED = 404; // Guaranteed resource changed QUEUE_APP = 405; // Application changed QUEUE_ALLOC = 406; // Allocation changed ALLOC_CANCEL = 500; // Allocation cancelled by the RM ALLOC_PREEMPT = 501; // Allocation preempted by the core ALLOC_TIMEOUT = 502; // Allocation cancelled due to timeout ALLOC_REPLACED = 503; // Allocation replacement (placeholder) ALLOC_NODEREMOVED = 504; // Allocation cancelled, node removal } // the type of the object associated with the event Type type = 1; // ID of the object associated with the event string objectID = 2; // the detailed message as string string message = 5; // timestamp of the event int64 timestampNano = 6; // the type of the change ChangeType eventChangeType = 7; // details about the change ChangeDetail eventChangeDetail = 8; // the secondary object in the event (eg. allocation UUID, request ID) string referenceID = 9; // the resource value if the change involves setting/modifying a resource Resource resource = 10; reserved 3; reserved "groupID"; reserved 4; reserved "reason"; }