Github user cloud-fan commented on a diff in the pull request: https://github.com/apache/spark/pull/22009#discussion_r208693454 --- Diff: sql/core/src/main/java/org/apache/spark/sql/sources/v2/reader/streaming/ContinuousReadSupport.java --- @@ -0,0 +1,79 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.spark.sql.sources.v2.reader.streaming; + +import org.apache.spark.annotation.InterfaceStability; +import org.apache.spark.sql.execution.streaming.BaseStreamingSource; +import org.apache.spark.sql.sources.v2.reader.InputPartition; +import org.apache.spark.sql.sources.v2.reader.ScanConfig; +import org.apache.spark.sql.sources.v2.reader.ScanConfigBuilder; + +/** + * An interface that defines how to scan the data from data source for continuous streaming + * processing. + * + * The execution engine will create an instance of this interface at the start of a streaming query, + * then call {@link #newScanConfigBuilder(Offset)} and create an instance of {@link ScanConfig} for + * the duration of the streaming query or until {@link #needsReconfiguration(ScanConfig)} is true. + * The {@link ScanConfig} will be used to create input partitions and reader factory to process data + * for its duration. At the end {@link #stop()} will be called when the streaming execution is + * completed. Note that a single query may have multiple executions due to restart or failure + * recovery. + */ +@InterfaceStability.Evolving +public interface ContinuousReadSupport extends StreamingReadSupport, BaseStreamingSource { + + /** + * Returns a builder of {@link ScanConfig}. The builder can take some query specific information + * to do operators pushdown, streaming offsets, etc., and keep these information in the + * created {@link ScanConfig}. + * + * This is the first step of the data scan. All other methods in {@link ContinuousReadSupport} + * needs to take {@link ScanConfig} as an input. + * + * If this method fails (by throwing an exception), the action will fail and no Spark job will be + * submitted. + */ + ScanConfigBuilder newScanConfigBuilder(Offset start); + + /** + * Returns a factory to produce {@link ContinuousPartitionReader}s for {@link InputPartition}s. + * + * If this method fails (by throwing an exception), the action will fail and no Spark job will be + * submitted. + */ + ContinuousPartitionReaderFactory createContinuousReaderFactory(ScanConfig config); + + /** + * Merge partitioned offsets coming from {@link ContinuousPartitionReader} instances + * for each partition to a single global offset. + */ + Offset mergeOffsets(PartitionOffset[] offsets); + + /** + * The execution engine will call this method in every epoch to determine if new input + * partitions need to be generated, which may be required if for example the underlying + * source system has had partitions added or removed. + * + * If true, the query will be shut down and restarted with a new {@link ContinuousReadSupport} --- End diff -- @jose-torres with the new abstraction of `ScanConfig`, do we still need to create a new `ContinuousReadSupport` for reconfiguration?
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org