Package org.apache.flink.api.connector.sink2

Interface WriterInitContext

All Superinterfaces:

InitContext

@Public
public interface WriterInitContext
extends InitContext

The interface exposes some runtime info for creating a SinkWriter.

Field Summary

Fields inherited from interface org.apache.flink.api.connector.sink2.InitContext

INITIAL_CHECKPOINT_ID

Method Summary

Modifier and Type	Method	Description
SerializationSchema.InitializationContext	asSerializationSchemaInitializationContext()	Provides a view on this context as a SerializationSchema.InitializationContext.
<IN> TypeSerializer<IN>	createInputSerializer()	Creates a serializer for the type of sink's input.
MailboxExecutor	getMailboxExecutor()	Returns the mailbox executor that allows to execute Runnables inside the task thread in between record processing.
ProcessingTimeService	getProcessingTimeService()	Returns a ProcessingTimeService that can be used to get the current time and register timers.
UserCodeClassLoader	getUserCodeClassLoader()	Gets the UserCodeClassLoader to load classes that are not in system's classpath, but are part of the jar file of a user job.
boolean	isObjectReuseEnabled()	Returns whether object reuse has been enabled or disabled.
default <MetaT> Optional<java.util.function.Consumer<MetaT>>	metadataConsumer()	Returns a metadata consumer, the SinkWriter can publish metadata events of type WriterInitContext to the consumer.
org.apache.flink.metrics.groups.SinkWriterMetricGroup	metricGroup()

Methods inherited from interface org.apache.flink.api.connector.sink2.InitContext

getAttemptNumber, getJobId, getJobInfo, getNumberOfParallelSubtasks, getRestoredCheckpointId, getSubtaskId, getTaskInfo

Method Detail

getUserCodeClassLoader

UserCodeClassLoader getUserCodeClassLoader()

Gets the UserCodeClassLoader to load classes that are not in system's classpath, but are part of the jar file of a user job.

See Also:

UserCodeClassLoader

getMailboxExecutor

MailboxExecutor getMailboxExecutor()

Returns the mailbox executor that allows to execute Runnables inside the task thread in between record processing.

Note that this method should not be used per-record for performance reasons in the same way as records should not be sent to the external system individually. Rather, implementers are expected to batch records and only enqueue a single Runnable per batch to handle the result.

getProcessingTimeService

ProcessingTimeService getProcessingTimeService()

Returns a ProcessingTimeService that can be used to get the current time and register timers.

metricGroup

org.apache.flink.metrics.groups.SinkWriterMetricGroup metricGroup()

Returns:

The metric group this writer belongs to.

asSerializationSchemaInitializationContext

SerializationSchema.InitializationContext asSerializationSchemaInitializationContext()

Provides a view on this context as a SerializationSchema.InitializationContext.

isObjectReuseEnabled

boolean isObjectReuseEnabled()

Returns whether object reuse has been enabled or disabled.

createInputSerializer

<IN> TypeSerializer<IN> createInputSerializer()

Creates a serializer for the type of sink's input.

metadataConsumer

default <MetaT> Optional<java.util.function.Consumer<MetaT>> metadataConsumer()

Returns a metadata consumer, the SinkWriter can publish metadata events of type WriterInitContext to the consumer.

It is recommended to use a separate thread pool to publish the metadata because enqueuing a lot of these messages in the mailbox may lead to a performance decrease. thread, and the Consumer.accept(T) method is executed very fast.