org.tensorflow.proto.data.experimental

Interface ServiceConfig.DispatcherConfigOrBuilder

All Superinterfaces:

com.google.protobuf.MessageLiteOrBuilder, com.google.protobuf.MessageOrBuilder

All Known Implementing Classes:

ServiceConfig.DispatcherConfig, ServiceConfig.DispatcherConfig.Builder

Enclosing class:

ServiceConfig

public static interface ServiceConfig.DispatcherConfigOrBuilder
extends com.google.protobuf.MessageOrBuilder

Method Summary

Modifier and Type	Method and Description
long	getClientTimeoutMs() How long to wait before garbage-collecting a client that hasn't heartbeated to the dispatcher.
boolean	getFaultTolerantMode() Whether to run in fault tolerant mode, where dispatcher state is saved across restarts.
long	getJobGcCheckIntervalMs() How often the dispatcher should scan through to delete old and unused jobs.
long	getJobGcTimeoutMs() How long a job needs to be unused before it becomes a candidate for garbage collection.
long	getPort() The port for the dispatcher to bind to.
String	getProtocol() The protocol for the dispatcher to use when connecting to workers.
com.google.protobuf.ByteString	getProtocolBytes() The protocol for the dispatcher to use when connecting to workers.
String	getWorkDir() A work directory to use for storing dispatcher state, and for recovering during restarts.
com.google.protobuf.ByteString	getWorkDirBytes() A work directory to use for storing dispatcher state, and for recovering during restarts.
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.
com.google.protobuf.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.
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.
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.

Methods inherited from interface com.google.protobuf.MessageOrBuilder

findInitializationErrors, getAllFields, getDefaultInstanceForType, getDescriptorForType, getField, getInitializationErrorString, getOneofFieldDescriptor, getRepeatedField, getRepeatedFieldCount, getUnknownFields, hasField, hasOneof

Methods inherited from interface com.google.protobuf.MessageLiteOrBuilder

isInitialized

Method Detail

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;

getProtocol

String getProtocol()

 The protocol for the dispatcher to use when connecting to workers.
 

string protocol = 2;

getProtocolBytes

com.google.protobuf.ByteString getProtocolBytes()

 The protocol for the dispatcher to use when connecting to workers.
 

string protocol = 2;

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;

getWorkDirBytes

com.google.protobuf.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;

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;

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.
 TODO(yangchen): Also add a deployment mode flag. Auto-sharding will only be
 allowed in the "COLOCATED" mode.
 

repeated string worker_addresses = 7;

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.
 TODO(yangchen): Also add a deployment mode flag. Auto-sharding will only be
 allowed in the "COLOCATED" mode.
 

repeated string worker_addresses = 7;

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.
 TODO(yangchen): Also add a deployment mode flag. Auto-sharding will only be
 allowed in the "COLOCATED" mode.
 

repeated string worker_addresses = 7;

getWorkerAddressesBytes

com.google.protobuf.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.
 TODO(yangchen): Also add a deployment mode flag. Auto-sharding will only be
 allowed in the "COLOCATED" mode.
 

repeated string worker_addresses = 7;

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;

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.
 

int64 job_gc_timeout_ms = 6;

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;