syntax = "proto3";

package tensorflow;

option go_package = "github.com/tensorflow/tensorflow/tensorflow/go/core/protobuf/for_core_protos_go_proto";

// Represents a job type and the number of tasks under this job.
// For example, ("worker", 20) implies that there will be 20 worker tasks.
message CoordinatedJob {
  string name = 1;
  int32 num_tasks = 2;
}

// Coordination service configuration parameters.
// The system picks appropriate values for fields that are not set.
message CoordinationServiceConfig {
  // Type of coordination service implementation to enable.
  // For example, setting the service type as "standalone" starts a service
  // instance on the leader task to provide the coordination services such as
  // heartbeats and consistent key-value store.
  string service_type = 1;

  // Address where the coordination service instance is hosted.
  string service_leader = 2;

  // Whether to enable the health check mechanism.
  bool enable_health_check = 3;

  // Maximum wait time for all members in the cluster to be registered.
  int64 cluster_register_timeout_in_ms = 4;

  // Denotes if we should synchronize the agents' register attempts by blocking
  // on a barrier. This is useful for synchronized restarts.
  bool cluster_register_with_barrier = 14;

  // Heartbeat timeout, if a task does not record heartbeat in this time
  // window, it will be considered disconnected.
  // Note: This is also used as a grace period to accept any heartbeats after
  // the agent has disconnected, to account for the lag time between the service
  // recording the state change and the agent stopping heartbeats.
  int64 heartbeat_timeout_in_ms = 5;

  // The list of `CoordinatedJob`s that will register in coordination service.
  reserved 6;
  repeated CoordinatedJob coordinated_job_list = 10;

  // Denotes how long to wait for all coordination agents to reach the barriers
  // (after the first shutdown request) before disconnecting together. If
  // set to 0, no barrier is imposed upon shutdown and each worker can
  // disconnect individually.
  int64 shutdown_barrier_timeout_in_ms = 7;

  // If set, agents do not make an explicit Shutdown() call. Service will only
  // find out about the disconnecte agent via stale heartbeats. Used for
  // testing.
  bool agent_destruction_without_shutdown = 8;

  // The list of jobs which are recoverable. If a task in this list fails,
  // it will not propagate error to other tasks.
  // If empty, no jobs will be recoverable and every task failure will cause
  // error propagation to other tasks.
  repeated string recoverable_jobs = 9;

  // If a task restarts with a new incarnation, we may allow it to reconnect
  // silently. This is useful when we know that a task can immediately resume
  // work upon re-connecting to the service.
  bool allow_new_incarnation_to_reconnect = 11;

  // Disables coordination service.
  // Some libraries enable coordination service by default even if the user did
  // not specify any config. This field allows users to explicitly disable
  // coordination service under all situations.
  bool force_disable = 12;

  // Use long polling to get error from coordination service as the error
  // propagation mechanism.
  bool poll_for_error_from_service_at_startup = 13;
}