com.google.cloud.dataflow.sdk.io

Class PubsubUnboundedSink<T>

java.lang.Object

com.google.cloud.dataflow.sdk.transforms.PTransform<PCollection<T>,PDone>

com.google.cloud.dataflow.sdk.io.PubsubUnboundedSink<T>

All Implemented Interfaces:

HasDisplayData, Serializable

public class PubsubUnboundedSink<T>
extends PTransform<PCollection<T>,PDone>

A PTransform which streams messages to Pubsub.

• The underlying implementation is just a GroupByKey followed by a ParDo which publishes as a side effect. (In the future we want to design and switch to a custom UnboundedSink implementation so as to gain access to system watermark and end-of-pipeline cleanup.)
• We try to send messages in batches while also limiting send latency.
• No stats are logged. Rather some counters are used to keep track of elements and batches.
• Though some background threads are used by the underlying netty system all actual Pubsub calls are blocking. We rely on the underlying runner to allow multiple DoFn instances to execute concurrently and hide latency.
• A failed bundle will cause messages to be resent. Thus we rely on the Pubsub consumer to dedup messages.

NOTE: This is not the implementation used when running on the Google Cloud Dataflow service.

See Also:

Serialized Form

Field Summary

Fields inherited from class com.google.cloud.dataflow.sdk.transforms.PTransform

name

Constructor Summary

Constructors

Constructor and Description
PubsubUnboundedSink(com.google.cloud.dataflow.sdk.util.PubsubClient.PubsubClientFactory pubsubFactory, com.google.cloud.dataflow.sdk.util.PubsubClient.TopicPath topic, Coder<T> elementCoder, String timestampLabel, String idLabel, int numShards)

Method Summary

Modifier and Type	Method and Description
PDone	apply(PCollection<T> input) Applies this PTransform on the given InputT, and returns its Output.
Coder<T>	getElementCoder()
String	getIdLabel()
String	getTimestampLabel()
com.google.cloud.dataflow.sdk.util.PubsubClient.TopicPath	getTopic()

Methods inherited from class com.google.cloud.dataflow.sdk.transforms.PTransform

getDefaultOutputCoder, getDefaultOutputCoder, getDefaultOutputCoder, getKindString, getName, populateDisplayData, toString, validate

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

PubsubUnboundedSink

public PubsubUnboundedSink(com.google.cloud.dataflow.sdk.util.PubsubClient.PubsubClientFactory pubsubFactory,
                           com.google.cloud.dataflow.sdk.util.PubsubClient.TopicPath topic,
                           Coder<T> elementCoder,
                           String timestampLabel,
                           String idLabel,
                           int numShards)

Method Detail

getTopic

public com.google.cloud.dataflow.sdk.util.PubsubClient.TopicPath getTopic()

getTimestampLabel

@Nullable
public String getTimestampLabel()

getIdLabel

@Nullable
public String getIdLabel()

getElementCoder

public Coder<T> getElementCoder()

apply

public PDone apply(PCollection<T> input)

Description copied from class: PTransform

Applies this PTransform on the given InputT, and returns its Output.

Composite transforms, which are defined in terms of other transforms, should return the output of one of the composed transforms. Non-composite transforms, which do not apply any transforms internally, should return a new unbound output and register evaluators (via backend-specific registration methods).

The default implementation throws an exception. A derived class must either implement apply, or else each runner must supply a custom implementation via PipelineRunner.apply(com.google.cloud.dataflow.sdk.transforms.PTransform<InputT, OutputT>, InputT).

Overrides:

apply in class PTransform<PCollection<T>,PDone>