Interface ServiceConfig.DispatcherConfigOrBuilder

All Superinterfaces:
MessageLiteOrBuilder, MessageOrBuilder
All Known Implementing Classes:
ServiceConfig.DispatcherConfig, ServiceConfig.DispatcherConfig.Builder
Enclosing class:
ServiceConfig
public static interface ServiceConfig.DispatcherConfigOrBuilder extends MessageOrBuilder

  • Method Details

    • getPort

      long getPort()
      The port for the dispatcher to bind to. A value of 0 indicates that the
dispatcher may bind to any available port.
      int64 port = 1;
      Returns:
      The port.

    • getProtocol

      String getProtocol()
      The protocol for the dispatcher to use when connecting to workers.
      string protocol = 2;
      Returns:
      The protocol.

    • getProtocolBytes

      ByteString getProtocolBytes()
      The protocol for the dispatcher to use when connecting to workers.
      string protocol = 2;
      Returns:
      The bytes for protocol.

    • getWorkDir

      String getWorkDir()
      A work directory to use for storing dispatcher state, and for recovering
during restarts. The empty string indicates not to use any work directory.
      string work_dir = 3;
      Returns:
      The workDir.

    • getWorkDirBytes

      ByteString getWorkDirBytes()
      A work directory to use for storing dispatcher state, and for recovering
during restarts. The empty string indicates not to use any work directory.
      string work_dir = 3;
      Returns:
      The bytes for workDir.

    • getFaultTolerantMode

      boolean getFaultTolerantMode()
      Whether to run in fault tolerant mode, where dispatcher state is saved
across restarts. Requires that `work_dir` is nonempty.
      bool fault_tolerant_mode = 4;
      Returns:
      The faultTolerantMode.

    • getWorkerAddressesList

      List<String> getWorkerAddressesList()
      (Optional.) If the job uses auto-sharding, it needs to specify a fixed list
of worker addresses that will register with the dispatcher. The worker
addresses should be in the format "host" or "host:port", where "port" is an
integer, named port, or %port% to match any port.
      repeated string worker_addresses = 7;
      Returns:
      A list containing the workerAddresses.

    • getWorkerAddressesCount

      int getWorkerAddressesCount()
      (Optional.) If the job uses auto-sharding, it needs to specify a fixed list
of worker addresses that will register with the dispatcher. The worker
addresses should be in the format "host" or "host:port", where "port" is an
integer, named port, or %port% to match any port.
      repeated string worker_addresses = 7;
      Returns:
      The count of workerAddresses.

    • getWorkerAddresses

      String getWorkerAddresses(int index)
      (Optional.) If the job uses auto-sharding, it needs to specify a fixed list
of worker addresses that will register with the dispatcher. The worker
addresses should be in the format "host" or "host:port", where "port" is an
integer, named port, or %port% to match any port.
      repeated string worker_addresses = 7;
      Parameters:
      index - The index of the element to return.
      Returns:
      The workerAddresses at the given index.

    • getWorkerAddressesBytes

      ByteString getWorkerAddressesBytes(int index)
      (Optional.) If the job uses auto-sharding, it needs to specify a fixed list
of worker addresses that will register with the dispatcher. The worker
addresses should be in the format "host" or "host:port", where "port" is an
integer, named port, or %port% to match any port.
      repeated string worker_addresses = 7;
      Parameters:
      index - The index of the value to return.
      Returns:
      The bytes of the workerAddresses at the given index.

    • getDeploymentModeValue

      int getDeploymentModeValue()
      (Optional.) tf.data service deployment mode. Supported values are "REMOTE",
"COLOCATED", and "HYBRID". If unspecified, it is assumed to be "REMOTE".
      .tensorflow.data.DeploymentMode deployment_mode = 9;
      Returns:
      The enum numeric value on the wire for deploymentMode.

    • getDeploymentMode

      DataService.DeploymentMode getDeploymentMode()
      (Optional.) tf.data service deployment mode. Supported values are "REMOTE",
"COLOCATED", and "HYBRID". If unspecified, it is assumed to be "REMOTE".
      .tensorflow.data.DeploymentMode deployment_mode = 9;
      Returns:
      The deploymentMode.

    • getJobGcCheckIntervalMs

      long getJobGcCheckIntervalMs()
      How often the dispatcher should scan through to delete old and unused
jobs. A value of 0 indicates that the decision should be left up to the
runtime.
      int64 job_gc_check_interval_ms = 5;
      Returns:
      The jobGcCheckIntervalMs.

    • getJobGcTimeoutMs

      long getJobGcTimeoutMs()
      How long a job needs to be unused before it becomes a candidate for garbage
collection. A value of -1 indicates that jobs should never be garbage
collected. A value of 0 indicates that the decision should be left up to
the runtime. Note: This does not apply to dynamic sharding unless users
explicitly opt-in by enabling `gc_dynamic_sharding_jobs` below.
      int64 job_gc_timeout_ms = 6;
      Returns:
      The jobGcTimeoutMs.

    • getGcDynamicShardingJobs

      boolean getGcDynamicShardingJobs()
      Whether dynamically sharded jobs should be eligible for garbage collection.
These jobs are not garbage collected by default, since if a job is garbage
collected and then re-created, it will revisit all data from the start. If
revisiting data is acceptible and you want automatic reclamation of
iterator memory, set `gc_dynamic_sharding_jobs` to `true`.
      bool gc_dynamic_sharding_jobs = 11;
      Returns:
      The gcDynamicShardingJobs.

    • getClientTimeoutMs

      long getClientTimeoutMs()
      How long to wait before garbage-collecting a client that hasn't
heartbeated to the dispatcher. A value of 0 indicates that the timeout
should be left to the runtime.
      int64 client_timeout_ms = 8;
      Returns:
      The clientTimeoutMs.

    • getWorkerTimeoutMs

      long getWorkerTimeoutMs()
      How long to wait for a worker to heartbeat before considering it missing.
A value of 0 indicates that the timeout should be left to the runtime.
      int64 worker_timeout_ms = 10;
      Returns:
      The workerTimeoutMs.

    • getWorkerMaxConcurrentSnapshots

      long getWorkerMaxConcurrentSnapshots()
      The maximum number of snapshots that a worker can concurrently process at a
given point in time. This is a tradeoff between worker resource usage and
snapshot wall time. A value of 0 indicates that the decision should be left
up to the runtime.
      int64 worker_max_concurrent_snapshots = 12;
      Returns:
      The workerMaxConcurrentSnapshots.