Field Detail

• ONE_KB

  static final int ONE_KB

  1KB in bytes

  See Also:

  Constant Field Values

• ONE_MB

  static final int ONE_MB

  1MB in bytes

  See Also:

  Constant Field Values

• clientCacheSize

  static final int clientCacheSize

  See Also:

  Constant Field Values

• NETTY_REQUEST_ENCODER_BUFFER_SIZE

  static final org.apache.giraph.conf.IntConfOption NETTY_REQUEST_ENCODER_BUFFER_SIZE

  How big to make the encoder buffer?

• CLIENT_SEND_BUFFER_SIZE

  static final org.apache.giraph.conf.IntConfOption CLIENT_SEND_BUFFER_SIZE

  Client send buffer size

• SERVER_RECEIVE_BUFFER_SIZE

  static final org.apache.giraph.conf.IntConfOption SERVER_RECEIVE_BUFFER_SIZE

  Server receive buffer size. a little bit larger than request size.

• MAX_FRAME_LENGTH

  static final org.apache.giraph.conf.IntConfOption MAX_FRAME_LENGTH

• MAX_OUT_MSG_CACHE_SIZE

  static final org.apache.giraph.conf.IntConfOption MAX_OUT_MSG_CACHE_SIZE

  Should be used by byteBuf cache.

• VERTEX_OUTPUT_FORMAT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.VertexOutputFormat> VERTEX_OUTPUT_FORMAT_CLASS

  VertexOutputFormat class

• VERTEX_INPUT_FORMAT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.VertexInputFormat> VERTEX_INPUT_FORMAT_CLASS

  VertexOutputFormat class

• EDGE_INPUT_FORMAT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.EdgeInputFormat> EDGE_INPUT_FORMAT_CLASS

  EdgeInputFormat class

• VERTEX_OUTPUT_FORMAT_SUBDIR

  static final org.apache.giraph.conf.StrConfOption VERTEX_OUTPUT_FORMAT_SUBDIR

  vertexOutputFormat sub-directory In Giraph, the output is to hdfs, they use the parent directory of default parent directory, and use sub directory as sub.

  In our project, we deem this configuration as ABSOLUTE path.

• VERTEX_OUTPUT_PATH

  static final org.apache.giraph.conf.StrConfOption VERTEX_OUTPUT_PATH

• VERTEX_ID_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.hadoop.io.WritableComparable> VERTEX_ID_CLASS

  Vertex index class

• VERTEX_VALUE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.hadoop.io.Writable> VERTEX_VALUE_CLASS

  Vertex value class

• EDGE_VALUE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.hadoop.io.Writable> EDGE_VALUE_CLASS

  Edge value class

• VERTEX_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.graph.Vertex> VERTEX_CLASS

  Vertex class

• OUTGOING_MESSAGE_VALUE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.hadoop.io.Writable> OUTGOING_MESSAGE_VALUE_CLASS

  Outgoing message value class

• INCOMING_MESSAGE_VALUE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.hadoop.io.Writable> INCOMING_MESSAGE_VALUE_CLASS

  incoming message value class

• WORKER_CONTEXT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<WorkerContext> WORKER_CONTEXT_CLASS

  Worker context class

• COMPUTATION_CLASS

  static final org.apache.giraph.conf.ClassConfOption<AbstractComputation> COMPUTATION_CLASS

  Worker context class

• TYPES_HOLDER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.conf.TypesHolder> TYPES_HOLDER_CLASS

  TypesHolder, used if Computation not set - optional

• MESSAGE_COMBINER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.combiner.MessageCombiner> MESSAGE_COMBINER_CLASS

  Message combiner class - optional

• OUTGOING_MESSAGE_VALUE_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.MessageValueFactory> OUTGOING_MESSAGE_VALUE_FACTORY_CLASS

  Outgoing message value factory class - optional

• MASTER_COMPUTE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<MasterCompute> MASTER_COMPUTE_CLASS

  Class for Master - optional

• CHANNELS_PER_SERVER

  static final org.apache.giraph.conf.IntConfOption CHANNELS_PER_SERVER

  Number of channels used per server

• REQUEST_SIZE_WARNING_THRESHOLD

  static final org.apache.giraph.conf.FloatConfOption REQUEST_SIZE_WARNING_THRESHOLD

  Warn if msg request size exceeds default size by this factor

• TCP_BACKLOG

  static final org.apache.giraph.conf.IntConfOption TCP_BACKLOG

  TCP backlog (defaults to number of workers)

• NETTY_USE_POOLED_ALLOCATOR

  static final org.apache.giraph.conf.BooleanConfOption NETTY_USE_POOLED_ALLOCATOR

  Use netty pooled memory buffer allocator

• NETTY_USE_DIRECT_MEMORY

  static final org.apache.giraph.conf.BooleanConfOption NETTY_USE_DIRECT_MEMORY

  Use direct memory buffers in netty

• CLIENT_RECEIVE_BUFFER_SIZE

  static final org.apache.giraph.conf.IntConfOption CLIENT_RECEIVE_BUFFER_SIZE

  Client receive buffer size

• SERVER_SEND_BUFFER_SIZE

  static final org.apache.giraph.conf.IntConfOption SERVER_SEND_BUFFER_SIZE

  Server send buffer size

• MESSAGE_AGGREGATE_SIZE

  static final org.apache.giraph.conf.IntConfOption MESSAGE_AGGREGATE_SIZE

  This represents the size of message cache, so actual cache bytes is generally (8 + 8) * aggregate_size; before flush.

  This is only used by batchWritable cache.

• NETTY_CLIENT_THREADS

  static final org.apache.giraph.conf.IntConfOption NETTY_CLIENT_THREADS

  Netty client threads

• NETTY_SERVER_BOSS_THREADS

  static final org.apache.giraph.conf.IntConfOption NETTY_SERVER_BOSS_THREADS

  Netty server boss threads

• NETTY_SERVER_WORKER_THREADS

  static final org.apache.giraph.conf.IntConfOption NETTY_SERVER_WORKER_THREADS

  Netty server worker threads

• NETTY_CLIENT_USE_EXECUTION_HANDLER

  static final org.apache.giraph.conf.BooleanConfOption NETTY_CLIENT_USE_EXECUTION_HANDLER

  Use the execution handler in netty on the client?

• NETTY_CLIENT_EXECUTION_THREADS

  static final org.apache.giraph.conf.IntConfOption NETTY_CLIENT_EXECUTION_THREADS

  Netty client execution threads (execution handler)

• NETTY_CLIENT_EXECUTION_AFTER_HANDLER

  static final org.apache.giraph.conf.StrConfOption NETTY_CLIENT_EXECUTION_AFTER_HANDLER

  Where to place the netty client execution handle?

• NETTY_SERVER_USE_EXECUTION_HANDLER

  static final org.apache.giraph.conf.BooleanConfOption NETTY_SERVER_USE_EXECUTION_HANDLER

  Use the execution handler in netty on the server?

• NETTY_SERVER_EXECUTION_THREADS

  static final org.apache.giraph.conf.IntConfOption NETTY_SERVER_EXECUTION_THREADS

  Netty server execution threads (execution handler)

• NETTY_SERVER_EXECUTION_AFTER_HANDLER

  static final org.apache.giraph.conf.StrConfOption NETTY_SERVER_EXECUTION_AFTER_HANDLER

  Where to place the netty server execution handle?

• NETTY_SIMULATE_FIRST_REQUEST_CLOSED

  static final org.apache.giraph.conf.BooleanConfOption NETTY_SIMULATE_FIRST_REQUEST_CLOSED

  Netty simulate a first request closed

• NETTY_SIMULATE_FIRST_RESPONSE_FAILED

  static final org.apache.giraph.conf.BooleanConfOption NETTY_SIMULATE_FIRST_RESPONSE_FAILED

  Netty simulate a first response failed

• NETTY_COMPRESSION_ALGORITHM

  static final org.apache.giraph.conf.StrConfOption NETTY_COMPRESSION_ALGORITHM

  Netty - set which compression to use

• WAITING_REQUEST_MSECS

  static final org.apache.giraph.conf.IntConfOption WAITING_REQUEST_MSECS

  Msecs to wait between waiting for all requests to finish

• MAX_WORKERS

  static final String MAX_WORKERS

  Maximum number of simultaneous worker tasks started by this job (int).

  See Also:

  Constant Field Values

• IPC_INITIAL_PORT

  static final org.apache.giraph.conf.IntConfOption IPC_INITIAL_PORT

  Initial port to start using for the IPC communication

• MAX_IPC_PORT_BIND_ATTEMPTS

  static final org.apache.giraph.conf.IntConfOption MAX_IPC_PORT_BIND_ATTEMPTS

  Maximum bind attempts for different IPC ports

• MAX_CONN_TRY_ATTEMPTS

  static final org.apache.giraph.conf.IntConfOption MAX_CONN_TRY_ATTEMPTS

  Maximum connections tries for client to connect to server

• MESSAGE_MANAGER_BASE_SERVER_PORT

  static final org.apache.giraph.conf.IntConfOption MESSAGE_MANAGER_BASE_SERVER_PORT

• AGGREGATOR_BASE_SERVER_PORT

  static final org.apache.giraph.conf.IntConfOption AGGREGATOR_BASE_SERVER_PORT

• MESSAGE_MANAGER_TYPE

  static final org.apache.giraph.conf.StrConfOption MESSAGE_MANAGER_TYPE

• INET_ADDRESS_MAX_RESOLVE_TIMES

  static final org.apache.giraph.conf.IntConfOption INET_ADDRESS_MAX_RESOLVE_TIMES

• USE_UNSAFE_SERIALIZATION

  static final org.apache.giraph.conf.BooleanConfOption USE_UNSAFE_SERIALIZATION

  Use unsafe serialization?

• MESSAGE_STORE_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<MessageStoreFactory> MESSAGE_STORE_FACTORY_CLASS

  Message Store Factory

• USE_PRIMITIVE_MESSAGE_STORE

  static final org.apache.giraph.conf.BooleanConfOption USE_PRIMITIVE_MESSAGE_STORE

• OUT_MESSAGE_CACHE_TYPE

  static final org.apache.giraph.conf.StrConfOption OUT_MESSAGE_CACHE_TYPE

  Default use byteBuf message cache.

• EDGE_MANAGER

  static final org.apache.giraph.conf.StrConfOption EDGE_MANAGER

• COMPUTATION_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.ComputationFactory> COMPUTATION_FACTORY_CLASS

• VERTEX_EDGES_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.edge.OutEdges> VERTEX_EDGES_CLASS

  Vertex edges class - optional

• INPUT_VERTEX_EDGES_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.edge.OutEdges> INPUT_VERTEX_EDGES_CLASS

• GRAPH_PARTITIONER_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.partition.GraphPartitionerFactory> GRAPH_PARTITIONER_FACTORY_CLASS

  Graph partitioner factory class - optional

• EDGE_OUTPUT_FORMAT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.EdgeOutputFormat> EDGE_OUTPUT_FORMAT_CLASS

• MAPPING_INPUT_FORMAT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.MappingInputFormat> MAPPING_INPUT_FORMAT_CLASS

  MappingInputFormat class

• AGGREGATOR_WRITER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.aggregators.AggregatorWriter> AGGREGATOR_WRITER_CLASS

• MESSAGE_ENCODE_AND_STORE_TYPE

  static final org.apache.giraph.conf.EnumConfOption<org.apache.giraph.comm.messages.MessageEncodeAndStoreType> MESSAGE_ENCODE_AND_STORE_TYPE

• VERTEX_RESOLVER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.graph.VertexResolver> VERTEX_RESOLVER_CLASS

• VERTEX_VALUE_COMBINER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.graph.VertexValueCombiner> VERTEX_VALUE_COMBINER_CLASS

• PARTITION_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.partition.Partition> PARTITION_CLASS

• EDGE_INPUT_FILTER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.filters.EdgeInputFilter> EDGE_INPUT_FILTER_CLASS

  EdgeInputFilter class

• VERTEX_INPUT_FILTER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.io.filters.VertexInputFilter> VERTEX_INPUT_FILTER_CLASS

  VertexInputFilter class

• MAPPING_STORE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.mapping.MappingStore> MAPPING_STORE_CLASS

  Mapping related information

• MAPPING_STORE_OPS_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.mapping.MappingStoreOps> MAPPING_STORE_OPS_CLASS

  Class to use for performing read operations on mapping store

• LB_MAPPINGSTORE_UPPER

  static final org.apache.giraph.conf.IntConfOption LB_MAPPINGSTORE_UPPER

  Upper value of LongByteMappingStore

• LB_MAPPINGSTORE_LOWER

  static final org.apache.giraph.conf.IntConfOption LB_MAPPINGSTORE_LOWER

  Lower value of LongByteMappingStore

• EDGE_TRANSLATION_CLASS

  static final org.apache.giraph.conf.ClassConfOption EDGE_TRANSLATION_CLASS

  Class used to conduct expensive edge translation during vertex input

• EDGE_STORE_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.edge.EdgeStoreFactory> EDGE_STORE_FACTORY_CLASS

  Edge Store Factory

• GRAPH_TYPE_LANGUAGES

  static final org.apache.giraph.conf.PerGraphTypeEnumConfOption<org.apache.giraph.graph.Language> GRAPH_TYPE_LANGUAGES

  Language user's graph types are implemented in

• GRAPH_TYPES_NEEDS_WRAPPERS

  static final org.apache.giraph.conf.PerGraphTypeBooleanConfOption GRAPH_TYPES_NEEDS_WRAPPERS

  Whether user graph types need wrappers

• VERTEX_ID_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.VertexIdFactory> VERTEX_ID_FACTORY_CLASS

  Vertex id factory class - optional

• VERTEX_VALUE_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.VertexValueFactory> VERTEX_VALUE_FACTORY_CLASS

  Vertex value factory class - optional

• EDGE_VALUE_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.EdgeValueFactory> EDGE_VALUE_FACTORY_CLASS

  Edge value factory class - optional

• VERTEX_EDGES_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.OutEdgesFactory> VERTEX_EDGES_FACTORY_CLASS

  OutEdges factory class - optional

• INPUT_VERTEX_EDGES_FACTORY_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.factories.OutEdgesFactory> INPUT_VERTEX_EDGES_FACTORY_CLASS

  OutEdges for input factory class - optional

• MASTER_OBSERVER_CLASSES

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.master.MasterObserver> MASTER_OBSERVER_CLASSES

  Classes for Master Observer - optional

• WORKER_OBSERVER_CLASSES

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.worker.WorkerObserver> WORKER_OBSERVER_CLASSES

  Classes for Worker Observer - optional

• MAPPER_OBSERVER_CLASSES

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.graph.MapperObserver> MAPPER_OBSERVER_CLASSES

  Classes for Mapper Observer - optional

• GC_OBSERVER_CLASSES

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.utils.GcObserver> GC_OBSERVER_CLASSES

  Classes for GC Observer - optional

• COMPUTATION_LANGUAGE

  static final org.apache.giraph.conf.EnumConfOption<org.apache.giraph.graph.Language> COMPUTATION_LANGUAGE

  Which language computation is implemented in

• RESOLVER_CREATE_VERTEX_ON_MSGS

  static final org.apache.giraph.conf.BooleanConfOption RESOLVER_CREATE_VERTEX_ON_MSGS

  Option of whether to create vertexes that did not exist before but received messages

• JOB_OBSERVER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.job.GiraphJobObserver> JOB_OBSERVER_CLASS

  Observer class to watch over job status - optional

• JOB_RETRY_CHECKER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.job.GiraphJobRetryChecker> JOB_RETRY_CHECKER_CLASS

  Observer class to watch over job status - optional

• MAX_ALLOWED_JOB_TIME_MS

  static final org.apache.giraph.conf.LongConfOption MAX_ALLOWED_JOB_TIME_MS

  Maximum allowed time for job to run after getting all resources before it will be killed, in milliseconds (-1 if it has no limit)

• EDGE_OUTPUT_FORMAT_SUBDIR

  static final org.apache.giraph.conf.StrConfOption EDGE_OUTPUT_FORMAT_SUBDIR

  EdgeOutputFormat sub-directory

• GIRAPH_TEXT_OUTPUT_FORMAT_SEPARATOR

  static final org.apache.giraph.conf.StrConfOption GIRAPH_TEXT_OUTPUT_FORMAT_SEPARATOR

  GiraphTextOutputFormat Separator

• GIRAPH_TEXT_OUTPUT_FORMAT_REVERSE

  static final org.apache.giraph.conf.BooleanConfOption GIRAPH_TEXT_OUTPUT_FORMAT_REVERSE

  Reverse values in the output

• DO_OUTPUT_DURING_COMPUTATION

  static final org.apache.giraph.conf.BooleanConfOption DO_OUTPUT_DURING_COMPUTATION

  If you use this option, instead of having saving vertices in the end of application, saveVertex will be called right after each vertex.compute() is called. NOTE: This feature doesn't work well with checkpointing - if you restart from a checkpoint you won't have any output from previous supersteps.

• VERTEX_OUTPUT_FORMAT_THREAD_SAFE

  static final org.apache.giraph.conf.BooleanConfOption VERTEX_OUTPUT_FORMAT_THREAD_SAFE

  Vertex output format thread-safe - if your VertexOutputFormat allows several vertexWriters to be created and written to in parallel, you should set this to true.

• NUM_OUTPUT_THREADS

  static final org.apache.giraph.conf.IntConfOption NUM_OUTPUT_THREADS

  Number of threads for writing output in the end of the application

• GIRAPH_YARN_LIBJARS

  static final org.apache.giraph.conf.StrConfOption GIRAPH_YARN_LIBJARS

  conf key for comma-separated list of jars to export to YARN workers

• GIRAPH_YARN_CONF_FILE

  static final String GIRAPH_YARN_CONF_FILE

  Name of the XML file that will export our Configuration to YARN workers

  See Also:

  Constant Field Values

• GIRAPH_YARN_TASK_HEAP_MB_DEFAULT

  static final int GIRAPH_YARN_TASK_HEAP_MB_DEFAULT

  Giraph default heap size for all tasks when running on YARN profile

  See Also:

  Constant Field Values

• GIRAPH_YARN_TASK_HEAP_MB

  static final org.apache.giraph.conf.IntConfOption GIRAPH_YARN_TASK_HEAP_MB

  Name of Giraph property for user-configurable heap memory per worker

• GIRAPH_YARN_PRIORITY

  static final int GIRAPH_YARN_PRIORITY

  Default priority level in YARN for our task containers

  See Also:

  Constant Field Values

• IS_PURE_YARN_JOB

  static final org.apache.giraph.conf.BooleanConfOption IS_PURE_YARN_JOB

  Is this a pure YARN job (i.e. no MapReduce layer managing Giraph tasks)

• MIN_WORKERS

  static final String MIN_WORKERS

  Minimum number of simultaneous workers before this job can run (int)

  See Also:

  Constant Field Values

• SPLIT_MASTER_WORKER

  static final org.apache.giraph.conf.BooleanConfOption SPLIT_MASTER_WORKER

  Separate the workers and the master tasks. This is required to support dynamic recovery. (boolean)

• LOCAL_TEST_MODE

  static final org.apache.giraph.conf.BooleanConfOption LOCAL_TEST_MODE

  Indicates whether this job is run in an internal unit test

• LOG_LEVEL

  static final org.apache.giraph.conf.StrConfOption LOG_LEVEL

  Override the Hadoop log level and set the desired log level.

• LOG_THREAD_LAYOUT

  static final org.apache.giraph.conf.BooleanConfOption LOG_THREAD_LAYOUT

  Use thread level debugging?

• JMAP_ENABLE

  static final org.apache.giraph.conf.BooleanConfOption JMAP_ENABLE

  Configuration key to enable jmap printing

• JMAP_SLEEP_MILLIS

  static final org.apache.giraph.conf.IntConfOption JMAP_SLEEP_MILLIS

  Configuration key for msec to sleep between calls

• JMAP_PRINT_LINES

  static final org.apache.giraph.conf.IntConfOption JMAP_PRINT_LINES

  Configuration key for how many lines to print

• JMAP_LIVE_ONLY

  static final org.apache.giraph.conf.BooleanConfOption JMAP_LIVE_ONLY

  Configuration key for printing live objects only This option will trigger Full GC for every jmap dump and so can significantly hinder performance.

• MIN_FREE_MBS_ON_HEAP

  static final org.apache.giraph.conf.IntConfOption MIN_FREE_MBS_ON_HEAP

  Option used by ReactiveJMapHistoDumper to check for an imminent OOM in worker or master process

• REACTIVE_JMAP_ENABLE

  static final org.apache.giraph.conf.BooleanConfOption REACTIVE_JMAP_ENABLE

  Option can be used to enable reactively dumping jmap histo when OOM is imminent

• MIN_PERCENT_RESPONDED

  static final org.apache.giraph.conf.FloatConfOption MIN_PERCENT_RESPONDED

  Minimum percent of the maximum number of workers that have responded in order to continue progressing. (float)

• METRICS_ENABLE

  static final org.apache.giraph.conf.BooleanConfOption METRICS_ENABLE

  Enable the Metrics system

• METRICS_DIRECTORY

  static final org.apache.giraph.conf.StrConfOption METRICS_DIRECTORY

  Directory in HDFS to write master metrics to, instead of stderr

• ZOOKEEPER_LIST

  static final org.apache.giraph.conf.StrConfOption ZOOKEEPER_LIST

  ZooKeeper comma-separated list (if not set, will start up ZooKeeper locally). Consider that after locally-starting zookeeper, this parameter will updated the configuration with the current configuration value.

• ZOOKEEPER_IS_EXTERNAL

  static final org.apache.giraph.conf.BooleanConfOption ZOOKEEPER_IS_EXTERNAL

  ZooKeeper List will always hold a value during the computation while this option provides information regarding whether the zookeeper was internally started or externally provided.

• ZOOKEEPER_SESSION_TIMEOUT

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_SESSION_TIMEOUT

  ZooKeeper session millisecond timeout

• ZOOKEEPER_SERVERLIST_POLL_MSECS

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_SERVERLIST_POLL_MSECS

  Polling interval to check for the ZooKeeper server data

• ZOOKEEPER_SERVER_PORT

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_SERVER_PORT

  ZooKeeper port to use

• ZOOKEEPER_DIR

  static final String ZOOKEEPER_DIR

  Local ZooKeeper directory to use

  See Also:

  Constant Field Values

• ZOOKEEPER_OPS_MAX_ATTEMPTS

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_OPS_MAX_ATTEMPTS

  Max attempts for handling ZooKeeper connection loss

• ZOOKEEPER_OPS_RETRY_WAIT_MSECS

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_OPS_RETRY_WAIT_MSECS

  Msecs to wait before retrying a failed ZooKeeper op due to connection loss.

• NETTY_SERVER_THREADS

  static final org.apache.giraph.conf.IntConfOption NETTY_SERVER_THREADS

  Netty server threads

• MAX_RESOLVE_ADDRESS_ATTEMPTS

  static final org.apache.giraph.conf.IntConfOption MAX_RESOLVE_ADDRESS_ATTEMPTS

  Max resolve address attempts

• EVENT_WAIT_MSECS

  static final org.apache.giraph.conf.IntConfOption EVENT_WAIT_MSECS

  Milliseconds to wait for an event before continuing

• MAX_MASTER_SUPERSTEP_WAIT_MSECS

  static final org.apache.giraph.conf.IntConfOption MAX_MASTER_SUPERSTEP_WAIT_MSECS

  Maximum milliseconds to wait before giving up trying to get the minimum number of workers before a superstep (int).

• MAX_COUNTER_WAIT_MSECS

  static final org.apache.giraph.conf.IntConfOption MAX_COUNTER_WAIT_MSECS

  Maximum milliseconds to wait before giving up waiting for the workers to write the counters to the ZooKeeper after a superstep

• MAX_REQUEST_MILLISECONDS

  static final org.apache.giraph.conf.IntConfOption MAX_REQUEST_MILLISECONDS

  Milliseconds for a request to complete (or else resend)

• RESEND_TIMED_OUT_REQUESTS

  static final org.apache.giraph.conf.BooleanConfOption RESEND_TIMED_OUT_REQUESTS

  Whether to resend request which timed out or fail the job if timeout happens

• NETTY_MAX_CONNECTION_FAILURES

  static final org.apache.giraph.conf.IntConfOption NETTY_MAX_CONNECTION_FAILURES

  Netty max connection failures

• WAIT_TIME_BETWEEN_CONNECTION_RETRIES_MS

  static final org.apache.giraph.conf.IntConfOption WAIT_TIME_BETWEEN_CONNECTION_RETRIES_MS

  How long to wait before trying to reconnect failed connections

• FAIL_FIRST_IPC_PORT_BIND_ATTEMPT

  static final org.apache.giraph.conf.BooleanConfOption FAIL_FIRST_IPC_PORT_BIND_ATTEMPT

  Fail first IPC port binding attempt, simulate binding failure on real grid testing

• MAX_MSG_REQUEST_SIZE

  static final org.apache.giraph.conf.IntConfOption MAX_MSG_REQUEST_SIZE

  Maximum size of messages (in bytes) per peer before flush

• ADDITIONAL_MSG_REQUEST_SIZE

  static final org.apache.giraph.conf.FloatConfOption ADDITIONAL_MSG_REQUEST_SIZE

  How much bigger than the average per partition size to make initial per partition buffers. If this value is A, message request size is M, and a worker has P partitions, than its initial partition buffer size will be (M / P) * (1 + A).

• MAX_VERTEX_REQUEST_SIZE

  static final org.apache.giraph.conf.IntConfOption MAX_VERTEX_REQUEST_SIZE

  Maximum size of vertices (in bytes) per peer before flush

• ADDITIONAL_VERTEX_REQUEST_SIZE

  static final org.apache.giraph.conf.FloatConfOption ADDITIONAL_VERTEX_REQUEST_SIZE

  Additional size (expressed as a ratio) of each per-partition buffer on top of the average size for vertices.

• MAX_EDGE_REQUEST_SIZE

  static final org.apache.giraph.conf.IntConfOption MAX_EDGE_REQUEST_SIZE

  Maximum size of edges (in bytes) per peer before flush

• ADDITIONAL_EDGE_REQUEST_SIZE

  static final org.apache.giraph.conf.FloatConfOption ADDITIONAL_EDGE_REQUEST_SIZE

  Additional size (expressed as a ratio) of each per-partition buffer on top of the average size for edges.

• MAX_MUTATIONS_PER_REQUEST

  static final org.apache.giraph.conf.IntConfOption MAX_MUTATIONS_PER_REQUEST

  Maximum number of mutations per partition before flush

• USE_MESSAGE_SIZE_ENCODING

  static final org.apache.giraph.conf.BooleanConfOption USE_MESSAGE_SIZE_ENCODING

  Use message size encoding (typically better for complex objects, not meant for primitive wrapped messages)

• MSG_NUM_FLUSH_THREADS

  static final String MSG_NUM_FLUSH_THREADS

  Number of flush threads per peer

  See Also:

  Constant Field Values

• NUM_COMPUTE_THREADS

  static final org.apache.giraph.conf.IntConfOption NUM_COMPUTE_THREADS

  Number of threads for vertex computation

• NUM_INPUT_THREADS

  static final org.apache.giraph.conf.IntConfOption NUM_INPUT_THREADS

  Number of threads for input split loading

• PARTITION_LONG_TAIL_MIN_PRINT

  static final org.apache.giraph.conf.IntConfOption PARTITION_LONG_TAIL_MIN_PRINT

  Minimum stragglers of the superstep before printing them out

• USE_SUPERSTEP_COUNTERS

  static final org.apache.giraph.conf.BooleanConfOption USE_SUPERSTEP_COUNTERS

  Use superstep counters? (boolean)

• INPUT_SPLIT_SAMPLE_PERCENT

  static final org.apache.giraph.conf.FloatConfOption INPUT_SPLIT_SAMPLE_PERCENT

  Input split sample percent - Used only for sampling and testing, rather than an actual job. The idea is that to test, you might only want a fraction of the actual input splits from your VertexInputFormat to load (values should be [0, 100]).

• INPUT_SPLIT_MAX_VERTICES

  static final org.apache.giraph.conf.LongConfOption INPUT_SPLIT_MAX_VERTICES

  To limit outlier vertex input splits from producing too many vertices or to help with testing, the number of vertices loaded from an input split can be limited. By default, everything is loaded.

• INPUT_SPLIT_MAX_EDGES

  static final org.apache.giraph.conf.LongConfOption INPUT_SPLIT_MAX_EDGES

  To limit outlier vertex input splits from producing too many vertices or to help with testing, the number of edges loaded from an input split can be limited. By default, everything is loaded.

• USE_INPUT_SPLIT_LOCALITY

  static final org.apache.giraph.conf.BooleanConfOption USE_INPUT_SPLIT_LOCALITY

  To minimize network usage when reading input splits, each worker can prioritize splits that reside on its host. This, however, comes at the cost of increased load on ZooKeeper. Hence, users with a lot of splits and input threads (or with configurations that can't exploit locality) may want to disable it.

• PARTITION_COUNT_MULTIPLIER

  static final org.apache.giraph.conf.FloatConfOption PARTITION_COUNT_MULTIPLIER

  Multiplier for the current workers squared

• MIN_PARTITIONS_PER_COMPUTE_THREAD

  static final org.apache.giraph.conf.IntConfOption MIN_PARTITIONS_PER_COMPUTE_THREAD

  Minimum number of partitions to have per compute thread

• USER_PARTITION_COUNT

  static final org.apache.giraph.conf.IntConfOption USER_PARTITION_COUNT

  Overrides default partition count calculation if not -1

• PARTITION_VERTEX_KEY_SPACE_SIZE

  static final String PARTITION_VERTEX_KEY_SPACE_SIZE

  Vertex key space size for WorkerGraphPartitionerImpl

  See Also:

  Constant Field Values

• CHECKPOINT_FREQUENCY

  static final org.apache.giraph.conf.IntConfOption CHECKPOINT_FREQUENCY

  How often to checkpoint (i.e. 0, means no checkpoint, 1 means every superstep, 2 is every two supersteps, etc.).

• CLEANUP_CHECKPOINTS_AFTER_SUCCESS

  static final org.apache.giraph.conf.BooleanConfOption CLEANUP_CHECKPOINTS_AFTER_SUCCESS

  Delete checkpoints after a successful job run?

• RESTART_SUPERSTEP

  static final String RESTART_SUPERSTEP

  An application can be restarted manually by selecting a superstep. The corresponding checkpoint must exist for this to work. The user should set a long value. Default is start from scratch.

  See Also:

  Constant Field Values

• RESTART_JOB_ID

  static final org.apache.giraph.conf.StrConfOption RESTART_JOB_ID

  If application is restarted manually we need to specify job ID to restart from.

• BASE_ZNODE_KEY

  static final String BASE_ZNODE_KEY

  Base ZNode for Giraph's state in the ZooKeeper cluster. Must be a root znode on the cluster beginning with "/"

  See Also:

  Constant Field Values

• ZOOKEEPER_MANAGER_DIRECTORY

  static final org.apache.giraph.conf.StrConfOption ZOOKEEPER_MANAGER_DIRECTORY

  If ZOOKEEPER_LIST is not set, then use this directory to manage ZooKeeper

• ZOOKEEPER_CONNECTION_ATTEMPTS

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_CONNECTION_ATTEMPTS

  Number of ZooKeeper client connection attempts before giving up.

• CHECKPOINT_DIRECTORY

  static final org.apache.giraph.conf.StrConfOption CHECKPOINT_DIRECTORY

  This directory has/stores the available checkpoint files in HDFS.

• PARTITIONS_DIRECTORY

  static final org.apache.giraph.conf.StrConfOption PARTITIONS_DIRECTORY

  Comma-separated list of directories in the local filesystem for out-of-core partitions.

• NUM_OUT_OF_CORE_THREADS

  static final org.apache.giraph.conf.IntConfOption NUM_OUT_OF_CORE_THREADS

  Number of IO threads used in out-of-core mechanism. If local disk is used for spilling data to and reading data from, this number should be equal to the number of available disks on each machine. In such case, one should use giraph.partitionsDirectory to specify directories mounted on different disks.

• USE_OUT_OF_CORE_GRAPH

  static final org.apache.giraph.conf.BooleanConfOption USE_OUT_OF_CORE_GRAPH

  Enable out-of-core graph.

• OUT_OF_CORE_DATA_ACCESSOR

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.ooc.persistence.OutOfCoreDataAccessor> OUT_OF_CORE_DATA_ACCESSOR

  Data accessor resource/object

• OUT_OF_CORE_ORACLE

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.ooc.policy.OutOfCoreOracle> OUT_OF_CORE_ORACLE

  Out-of-core oracle that is to be used for adaptive out-of-core engine. If the `MAX_PARTITIONS_IN_MEMORY` is already set, this will be over-written to be `FixedPartitionsOracle`.

• MAX_PARTITIONS_IN_MEMORY

  static final org.apache.giraph.conf.IntConfOption MAX_PARTITIONS_IN_MEMORY

  Maximum number of partitions to hold in memory for each worker.

• YOURKIT_OUTPUT_DIR

  static final String YOURKIT_OUTPUT_DIR

  Directory to write YourKit snapshots to

  See Also:

  Constant Field Values

• YOURKIT_OUTPUT_DIR_DEFAULT

  static final String YOURKIT_OUTPUT_DIR_DEFAULT

  Default directory to write YourKit snapshots to

  See Also:

  Constant Field Values

• KEEP_ZOOKEEPER_DATA

  static final org.apache.giraph.conf.BooleanConfOption KEEP_ZOOKEEPER_DATA

  Keep the zookeeper output for debugging? Default is to remove it.

• DEFAULT_ZOOKEEPER_SNAP_COUNT

  static final int DEFAULT_ZOOKEEPER_SNAP_COUNT

  Default ZooKeeper snap count.

  See Also:

  Constant Field Values

• DEFAULT_ZOOKEEPER_TICK_TIME

  static final int DEFAULT_ZOOKEEPER_TICK_TIME

  Default ZooKeeper tick time.

  See Also:

  Constant Field Values

• DEFAULT_ZOOKEEPER_MAX_CLIENT_CNXNS

  static final int DEFAULT_ZOOKEEPER_MAX_CLIENT_CNXNS

  Default ZooKeeper maximum client connections.

  See Also:

  Constant Field Values

• ZOOKEEPER_SNAP_RETAIN_COUNT

  static final int ZOOKEEPER_SNAP_RETAIN_COUNT

  Number of snapshots to be retained after purge

  See Also:

  Constant Field Values

• ZOOKEEPER_PURGE_INTERVAL

  static final int ZOOKEEPER_PURGE_INTERVAL

  ZooKeeper purge interval in hours

  See Also:

  Constant Field Values

• ZOOKEEPER_MIN_SESSION_TIMEOUT

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_MIN_SESSION_TIMEOUT

  ZooKeeper minimum session timeout

• ZOOKEEPER_MAX_SESSION_TIMEOUT

  static final org.apache.giraph.conf.IntConfOption ZOOKEEPER_MAX_SESSION_TIMEOUT

  ZooKeeper maximum session timeout

• ZOOKEEPER_FORCE_SYNC

  static final org.apache.giraph.conf.BooleanConfOption ZOOKEEPER_FORCE_SYNC

  ZooKeeper force sync

• ZOOKEEPER_SKIP_ACL

  static final org.apache.giraph.conf.BooleanConfOption ZOOKEEPER_SKIP_ACL

  ZooKeeper skip ACLs

• AUTHENTICATE

  static final org.apache.giraph.conf.BooleanConfOption AUTHENTICATE

  Whether to use SASL with DIGEST and Hadoop Job Tokens to authenticate and authorize Netty BSP Clients to Servers.

• SSL_ENCRYPT

  static final org.apache.giraph.conf.BooleanConfOption SSL_ENCRYPT

  Whether to use SSL to authenticate and authorize " Netty BSP Clients to Servers.

• USE_BIG_DATA_IO_FOR_MESSAGES

  static final org.apache.giraph.conf.BooleanConfOption USE_BIG_DATA_IO_FOR_MESSAGES

  Use BigDataIO for messages? If there are super-vertices in the graph which receive a lot of messages (total serialized size of messages goes beyond the maximum size of a byte array), setting this option to true will remove that limit. The maximum memory available for a single vertex will be limited to the maximum heap size available.

• MAX_TASK_ATTEMPTS

  static final org.apache.giraph.conf.IntConfOption MAX_TASK_ATTEMPTS

  Maximum number of attempts a master/worker will retry before killing the job. This directly maps to the number of map task attempts in Hadoop.

• DNS_INTERFACE

  static final org.apache.giraph.conf.StrConfOption DNS_INTERFACE

  Interface to use for hostname resolution

• DNS_NAMESERVER

  static final org.apache.giraph.conf.StrConfOption DNS_NAMESERVER

  Server for hostname resolution

• MAX_NUMBER_OF_SUPERSTEPS

  static final org.apache.giraph.conf.IntConfOption MAX_NUMBER_OF_SUPERSTEPS

  The application will halt after this many supersteps is completed. For instance, if it is set to 3, the application will run at most 0, 1, and 2 supersteps and then go into the shutdown superstep.

• STATIC_GRAPH

  static final org.apache.giraph.conf.BooleanConfOption STATIC_GRAPH

  The application will not mutate the graph topology (the edges). It is used to optimise out-of-core graph, by not writing back edges every time.

• CREATE_EDGE_SOURCE_VERTICES

  static final org.apache.giraph.conf.BooleanConfOption CREATE_EDGE_SOURCE_VERTICES

  This option can be used to specify if a source vertex present in edge input but not in vertex input can be created

• CREATE_EDGE_SOURCE_VERTICES_CALLBACK

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.edge.CreateSourceVertexCallback> CREATE_EDGE_SOURCE_VERTICES_CALLBACK

  Defines a call back that can be used to make decisions on whether the vertex should be created or not in the runtime.

• ZOOKEEPER_SERVER_PORT_COUNTER_GROUP

  static final String ZOOKEEPER_SERVER_PORT_COUNTER_GROUP

  This counter group will contain one counter whose name is the ZooKeeper server:port which this job is using

  See Also:

  Constant Field Values

• ZOOKEEPER_HALT_NODE_COUNTER_GROUP

  static final String ZOOKEEPER_HALT_NODE_COUNTER_GROUP

  This counter group will contain one counter whose name is the ZooKeeper node path which should be created to trigger computation halt

  See Also:

  Constant Field Values

• ZOOKEEPER_BASE_PATH_COUNTER_GROUP

  static final String ZOOKEEPER_BASE_PATH_COUNTER_GROUP

  This counter group will contain one counter whose name is the ZooKeeper node path which contains all data about this job

  See Also:

  Constant Field Values

• HALT_INSTRUCTIONS_WRITER_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.job.HaltApplicationUtils.HaltInstructionsWriter> HALT_INSTRUCTIONS_WRITER_CLASS

  Which class to use to write instructions on how to halt the application

• WAIT_TASK_DONE_TIMEOUT_MS

  static final org.apache.giraph.conf.IntConfOption WAIT_TASK_DONE_TIMEOUT_MS

  Maximum timeout (in milliseconds) for waiting for all tasks to complete after the job is done. Defaults to 15 minutes.

• TRACK_JOB_PROGRESS_ON_CLIENT

  static final org.apache.giraph.conf.BooleanConfOption TRACK_JOB_PROGRESS_ON_CLIENT

  Whether to track job progress on client or not

• JOB_PROGRESS_TRACKER_CLIENT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.graph.JobProgressTrackerClient> JOB_PROGRESS_TRACKER_CLIENT_CLASS

  Class to use as the job progress client

• JOB_PROGRESS_TRACKER_SERVICE_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.job.JobProgressTrackerService> JOB_PROGRESS_TRACKER_SERVICE_CLASS

  Class to use to track job progress on client

• VERTICES_TO_UPDATE_PROGRESS

  static final org.apache.giraph.conf.LongConfOption VERTICES_TO_UPDATE_PROGRESS

  Minimum number of vertices to compute before adding to worker progress.

• HDFS_FILE_CREATION_RETRIES

  static final org.apache.giraph.conf.IntConfOption HDFS_FILE_CREATION_RETRIES

  Number of retries for creating the HDFS files

• HDFS_FILE_CREATION_RETRY_WAIT_MS

  static final org.apache.giraph.conf.IntConfOption HDFS_FILE_CREATION_RETRY_WAIT_MS

  Number of milliseconds to wait before retrying HDFS file creation

• NUM_CHECKPOINT_IO_THREADS

  static final org.apache.giraph.conf.IntConfOption NUM_CHECKPOINT_IO_THREADS

  Number of threads for writing and reading checkpoints

• CHECKPOINT_COMPRESSION_CODEC

  static final org.apache.giraph.conf.StrConfOption CHECKPOINT_COMPRESSION_CODEC

  Compression algorithm to be used for checkpointing. Defined by extension for hadoop compatibility reasons.

• CHECKPOINT_SUPPORTED_CHECKER

  static final org.apache.giraph.conf.ClassConfOption<org.apache.giraph.bsp.checkpoints.CheckpointSupportedChecker> CHECKPOINT_SUPPORTED_CHECKER

  Defines if and when checkpointing is supported by this job. By default checkpointing is always supported unless output during the computation is enabled.

• ASYNC_MESSAGE_STORE_THREADS_COUNT

  static final org.apache.giraph.conf.IntConfOption ASYNC_MESSAGE_STORE_THREADS_COUNT

  Number of threads to use in async message store, 0 means we should not use async message processing

• HADOOP_OUTPUT_FORMAT_CLASS

  static final org.apache.giraph.conf.ClassConfOption<org.apache.hadoop.mapreduce.OutputFormat> HADOOP_OUTPUT_FORMAT_CLASS

  Output format class for hadoop to use (for committing)

• PREFER_IP_ADDRESSES

  static final org.apache.giraph.conf.BooleanConfOption PREFER_IP_ADDRESSES

  For worker to worker communication we can use IPs or host names, by default prefer IPs.

• WAIT_ZOOKEEPER_TIMEOUT_MSEC

  static final org.apache.giraph.conf.LongConfOption WAIT_ZOOKEEPER_TIMEOUT_MSEC

  Timeout for "waitForever", when we need to wait for zookeeper. Since we should never really have to wait forever. We should only wait some reasonable but large amount of time.

• WAIT_FOR_OTHER_WORKERS_TIMEOUT_MSEC

  static final org.apache.giraph.conf.LongConfOption WAIT_FOR_OTHER_WORKERS_TIMEOUT_MSEC

  Timeout for "waitForever", when we need to wait for other workers to complete their job. Since we should never really have to wait forever. We should only wait some reasonable but large amount of time.

• SUPERSTEP_COUNT

  static final org.apache.giraph.conf.IntConfOption SUPERSTEP_COUNT

  Number of supersteps job will run for

• DISABLE_GIRAPH_CLASS_RESOLVER

  static final org.apache.giraph.conf.BooleanConfOption DISABLE_GIRAPH_CLASS_RESOLVER

  Whether to disable GiraphClassResolver which is an efficient implementation of kryo class resolver. By default this resolver is used by KryoSimpleWritable and KryoSimpleWrapper, and can be disabled with this option

• JMAP_PATH

  static final org.apache.giraph.conf.StrConfOption JMAP_PATH

  Path where jmap exists

• FAIL_ON_EMPTY_INPUT

  static final org.apache.giraph.conf.BooleanConfOption FAIL_ON_EMPTY_INPUT

  Whether to fail the job or just warn when input is empty