SreeramGarlapati commented on a change in pull request #2660: URL: https://github.com/apache/iceberg/pull/2660#discussion_r653979695
########## File path: spark3/src/main/java/org/apache/iceberg/spark/source/SparkMicroBatchStream.java ########## @@ -0,0 +1,314 @@ +/* + * 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.iceberg.spark.source; + +import java.io.BufferedWriter; +import java.io.IOException; +import java.io.InputStream; +import java.io.OutputStream; +import java.io.OutputStreamWriter; +import java.nio.charset.StandardCharsets; +import java.util.ArrayList; +import java.util.List; +import org.apache.hadoop.fs.Path; +import org.apache.iceberg.CombinedScanTask; +import org.apache.iceberg.DataOperations; +import org.apache.iceberg.FileScanTask; +import org.apache.iceberg.MicroBatches; +import org.apache.iceberg.MicroBatches.MicroBatch; +import org.apache.iceberg.Schema; +import org.apache.iceberg.SchemaParser; +import org.apache.iceberg.SerializableTable; +import org.apache.iceberg.Snapshot; +import org.apache.iceberg.SnapshotSummary; +import org.apache.iceberg.Table; +import org.apache.iceberg.io.CloseableIterable; +import org.apache.iceberg.io.FileIO; +import org.apache.iceberg.io.InputFile; +import org.apache.iceberg.io.OutputFile; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.spark.Spark3Util; +import org.apache.iceberg.spark.SparkReadOptions; +import org.apache.iceberg.spark.source.SparkBatchScan.ReadTask; +import org.apache.iceberg.spark.source.SparkBatchScan.ReaderFactory; +import org.apache.iceberg.util.PropertyUtil; +import org.apache.iceberg.util.SnapshotUtil; +import org.apache.iceberg.util.TableScanUtil; +import org.apache.spark.api.java.JavaSparkContext; +import org.apache.spark.broadcast.Broadcast; +import org.apache.spark.sql.connector.read.InputPartition; +import org.apache.spark.sql.connector.read.PartitionReaderFactory; +import org.apache.spark.sql.connector.read.streaming.MicroBatchStream; +import org.apache.spark.sql.connector.read.streaming.Offset; +import org.apache.spark.sql.util.CaseInsensitiveStringMap; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import static org.apache.iceberg.TableProperties.SPLIT_LOOKBACK; +import static org.apache.iceberg.TableProperties.SPLIT_LOOKBACK_DEFAULT; +import static org.apache.iceberg.TableProperties.SPLIT_OPEN_FILE_COST; +import static org.apache.iceberg.TableProperties.SPLIT_OPEN_FILE_COST_DEFAULT; +import static org.apache.iceberg.TableProperties.SPLIT_SIZE; +import static org.apache.iceberg.TableProperties.SPLIT_SIZE_DEFAULT; + +public class SparkMicroBatchStream implements MicroBatchStream { + private static final Logger LOG = LoggerFactory.getLogger(SparkMicroBatchStream.class); + + private final JavaSparkContext sparkContext; + private final Table table; + private final boolean caseSensitive; + private final Schema expectedSchema; + private final Long splitSize; + private final Integer splitLookback; + private final Long splitOpenFileCost; + private final boolean localityPreferred; + private final StreamingOffset initialOffset; + + SparkMicroBatchStream(JavaSparkContext sparkContext, + Table table, boolean caseSensitive, Schema expectedSchema, + CaseInsensitiveStringMap options, String checkpointLocation) { + this.sparkContext = sparkContext; + this.table = table; + this.caseSensitive = caseSensitive; + this.expectedSchema = expectedSchema; + this.localityPreferred = Spark3Util.isLocalityEnabled(table.io(), table.location(), options); + + long tableSplitSize = PropertyUtil.propertyAsLong(table.properties(), SPLIT_SIZE, SPLIT_SIZE_DEFAULT); + this.splitSize = Spark3Util.propertyAsLong(options, SparkReadOptions.SPLIT_SIZE, tableSplitSize); + + int tableSplitLookback = PropertyUtil.propertyAsInt(table.properties(), SPLIT_LOOKBACK, SPLIT_LOOKBACK_DEFAULT); + this.splitLookback = Spark3Util.propertyAsInt(options, SparkReadOptions.LOOKBACK, tableSplitLookback); + + long tableSplitOpenFileCost = PropertyUtil.propertyAsLong( + table.properties(), SPLIT_OPEN_FILE_COST, SPLIT_OPEN_FILE_COST_DEFAULT); + this.splitOpenFileCost = Spark3Util.propertyAsLong(options, SPLIT_OPEN_FILE_COST, tableSplitOpenFileCost); + + InitialOffsetStore initialOffsetStore = InitialOffsetStore.getInstance(table, checkpointLocation); + this.initialOffset = getOrWriteInitialOffset(initialOffsetStore); + } + + @Override + public Offset latestOffset() { Review comment: based on our offline discussion with @rdblue - we landed on changing the default behavior to NOT `scanAllFiles` in first snapshot. I will introduce an option - if ppl want to start the stream with full scan. Flag name: `stream-start-mode = full | incremental` ########## File path: spark3/src/main/java/org/apache/iceberg/spark/source/SparkMicroBatchStream.java ########## @@ -0,0 +1,310 @@ +/* + * 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.iceberg.spark.source; + +import java.io.BufferedWriter; +import java.io.IOException; +import java.io.InputStream; +import java.io.OutputStream; +import java.io.OutputStreamWriter; +import java.nio.charset.StandardCharsets; +import java.util.ArrayList; +import java.util.List; +import org.apache.hadoop.fs.Path; +import org.apache.iceberg.CombinedScanTask; +import org.apache.iceberg.DataOperations; +import org.apache.iceberg.FileScanTask; +import org.apache.iceberg.MicroBatches; +import org.apache.iceberg.MicroBatches.MicroBatch; +import org.apache.iceberg.Schema; +import org.apache.iceberg.SchemaParser; +import org.apache.iceberg.SerializableTable; +import org.apache.iceberg.Snapshot; +import org.apache.iceberg.Table; +import org.apache.iceberg.io.CloseableIterable; +import org.apache.iceberg.io.FileIO; +import org.apache.iceberg.io.InputFile; +import org.apache.iceberg.io.OutputFile; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.relocated.com.google.common.collect.Iterables; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.spark.Spark3Util; +import org.apache.iceberg.spark.SparkReadOptions; +import org.apache.iceberg.spark.source.SparkBatchScan.ReadTask; +import org.apache.iceberg.spark.source.SparkBatchScan.ReaderFactory; +import org.apache.iceberg.util.PropertyUtil; +import org.apache.iceberg.util.SnapshotUtil; +import org.apache.iceberg.util.TableScanUtil; +import org.apache.spark.api.java.JavaSparkContext; +import org.apache.spark.broadcast.Broadcast; +import org.apache.spark.sql.connector.read.InputPartition; +import org.apache.spark.sql.connector.read.PartitionReaderFactory; +import org.apache.spark.sql.connector.read.streaming.MicroBatchStream; +import org.apache.spark.sql.connector.read.streaming.Offset; +import org.apache.spark.sql.util.CaseInsensitiveStringMap; + +import static org.apache.iceberg.TableProperties.SPLIT_LOOKBACK; +import static org.apache.iceberg.TableProperties.SPLIT_LOOKBACK_DEFAULT; +import static org.apache.iceberg.TableProperties.SPLIT_OPEN_FILE_COST; +import static org.apache.iceberg.TableProperties.SPLIT_OPEN_FILE_COST_DEFAULT; +import static org.apache.iceberg.TableProperties.SPLIT_SIZE; +import static org.apache.iceberg.TableProperties.SPLIT_SIZE_DEFAULT; + +public class SparkMicroBatchStream implements MicroBatchStream { + private final JavaSparkContext sparkContext; + private final Table table; + private final boolean caseSensitive; + private final Schema expectedSchema; + private final Long splitSize; + private final Integer splitLookback; + private final Long splitOpenFileCost; + private final boolean localityPreferred; + private final StreamingOffset initialOffset; + + SparkMicroBatchStream(JavaSparkContext sparkContext, + Table table, boolean caseSensitive, Schema expectedSchema, + CaseInsensitiveStringMap options, String checkpointLocation) { + this.sparkContext = sparkContext; + this.table = table; + this.caseSensitive = caseSensitive; + this.expectedSchema = expectedSchema; + this.localityPreferred = Spark3Util.isLocalityEnabled(table.io(), table.location(), options); + + long tableSplitSize = PropertyUtil.propertyAsLong(table.properties(), SPLIT_SIZE, SPLIT_SIZE_DEFAULT); + this.splitSize = Spark3Util.propertyAsLong(options, SparkReadOptions.SPLIT_SIZE, tableSplitSize); + + int tableSplitLookback = PropertyUtil.propertyAsInt(table.properties(), SPLIT_LOOKBACK, SPLIT_LOOKBACK_DEFAULT); + this.splitLookback = Spark3Util.propertyAsInt(options, SparkReadOptions.LOOKBACK, tableSplitLookback); + + long tableSplitOpenFileCost = PropertyUtil.propertyAsLong( + table.properties(), SPLIT_OPEN_FILE_COST, SPLIT_OPEN_FILE_COST_DEFAULT); + this.splitOpenFileCost = Spark3Util.propertyAsLong(options, SPLIT_OPEN_FILE_COST, tableSplitOpenFileCost); + + InitialOffsetStore initialOffsetStore = InitialOffsetStore.getInstance(table, checkpointLocation); + this.initialOffset = getOrWriteInitialOffset(initialOffsetStore); + } + + @Override + public Offset latestOffset() { + table.refresh(); + Snapshot latestSnapshot = table.currentSnapshot(); + if (latestSnapshot == null) { + return StreamingOffset.START_OFFSET; + } + + // a readStream on an Iceberg table can be started from 2 types of snapshots + // 1. a valid starting Snapshot: + // when this valid starting Snapshot is the initialOffset - then, scanAllFiles must be set to true; + // for all StreamingOffsets following this - scanAllFiles must be set to false + // 2. START_OFFSET: + // if the stream started on the table from START_OFFSET - it implies - that all the subsequent Snapshots added + // will have all files as net New manifests & hence scanAllFiles can be false. + boolean scanAllFiles = !StreamingOffset.START_OFFSET.equals(initialOffset) && + latestSnapshot.snapshotId() == initialOffset.snapshotId(); + + long filesNewlyAddedInLatestSnapshot = Iterables.size(latestSnapshot.addedFiles()); + long existingFilesInheritedByLatestSnapshot = SnapshotUtil.existingDataFiles(latestSnapshot); Review comment: based on my offline disc. with @rdblue - we decided to keep the original approach. Rationale: * current enumeration of existingFiles is doing very less work - loading that ManifestList file and reading the manifest stats. * `Snapshot.summary()` is not part of the Spec. So, any open - non-proprietary - implementations might not emit this. ########## File path: spark3/src/main/java/org/apache/iceberg/spark/source/SparkMicroBatchStream.java ########## @@ -0,0 +1,314 @@ +/* + * 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.iceberg.spark.source; + +import java.io.BufferedWriter; +import java.io.IOException; +import java.io.InputStream; +import java.io.OutputStream; +import java.io.OutputStreamWriter; +import java.nio.charset.StandardCharsets; +import java.util.ArrayList; +import java.util.List; +import org.apache.hadoop.fs.Path; +import org.apache.iceberg.CombinedScanTask; +import org.apache.iceberg.DataOperations; +import org.apache.iceberg.FileScanTask; +import org.apache.iceberg.MicroBatches; +import org.apache.iceberg.MicroBatches.MicroBatch; +import org.apache.iceberg.Schema; +import org.apache.iceberg.SchemaParser; +import org.apache.iceberg.SerializableTable; +import org.apache.iceberg.Snapshot; +import org.apache.iceberg.SnapshotSummary; +import org.apache.iceberg.Table; +import org.apache.iceberg.io.CloseableIterable; +import org.apache.iceberg.io.FileIO; +import org.apache.iceberg.io.InputFile; +import org.apache.iceberg.io.OutputFile; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.spark.Spark3Util; +import org.apache.iceberg.spark.SparkReadOptions; +import org.apache.iceberg.spark.source.SparkBatchScan.ReadTask; +import org.apache.iceberg.spark.source.SparkBatchScan.ReaderFactory; +import org.apache.iceberg.util.PropertyUtil; +import org.apache.iceberg.util.SnapshotUtil; +import org.apache.iceberg.util.TableScanUtil; +import org.apache.spark.api.java.JavaSparkContext; +import org.apache.spark.broadcast.Broadcast; +import org.apache.spark.sql.connector.read.InputPartition; +import org.apache.spark.sql.connector.read.PartitionReaderFactory; +import org.apache.spark.sql.connector.read.streaming.MicroBatchStream; +import org.apache.spark.sql.connector.read.streaming.Offset; +import org.apache.spark.sql.util.CaseInsensitiveStringMap; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import static org.apache.iceberg.TableProperties.SPLIT_LOOKBACK; +import static org.apache.iceberg.TableProperties.SPLIT_LOOKBACK_DEFAULT; +import static org.apache.iceberg.TableProperties.SPLIT_OPEN_FILE_COST; +import static org.apache.iceberg.TableProperties.SPLIT_OPEN_FILE_COST_DEFAULT; +import static org.apache.iceberg.TableProperties.SPLIT_SIZE; +import static org.apache.iceberg.TableProperties.SPLIT_SIZE_DEFAULT; + +public class SparkMicroBatchStream implements MicroBatchStream { + private static final Logger LOG = LoggerFactory.getLogger(SparkMicroBatchStream.class); + + private final JavaSparkContext sparkContext; + private final Table table; + private final boolean caseSensitive; + private final Schema expectedSchema; + private final Long splitSize; + private final Integer splitLookback; + private final Long splitOpenFileCost; + private final boolean localityPreferred; + private final StreamingOffset initialOffset; + + SparkMicroBatchStream(JavaSparkContext sparkContext, + Table table, boolean caseSensitive, Schema expectedSchema, + CaseInsensitiveStringMap options, String checkpointLocation) { + this.sparkContext = sparkContext; + this.table = table; + this.caseSensitive = caseSensitive; + this.expectedSchema = expectedSchema; + this.localityPreferred = Spark3Util.isLocalityEnabled(table.io(), table.location(), options); + + long tableSplitSize = PropertyUtil.propertyAsLong(table.properties(), SPLIT_SIZE, SPLIT_SIZE_DEFAULT); + this.splitSize = Spark3Util.propertyAsLong(options, SparkReadOptions.SPLIT_SIZE, tableSplitSize); + + int tableSplitLookback = PropertyUtil.propertyAsInt(table.properties(), SPLIT_LOOKBACK, SPLIT_LOOKBACK_DEFAULT); + this.splitLookback = Spark3Util.propertyAsInt(options, SparkReadOptions.LOOKBACK, tableSplitLookback); + + long tableSplitOpenFileCost = PropertyUtil.propertyAsLong( + table.properties(), SPLIT_OPEN_FILE_COST, SPLIT_OPEN_FILE_COST_DEFAULT); + this.splitOpenFileCost = Spark3Util.propertyAsLong(options, SPLIT_OPEN_FILE_COST, tableSplitOpenFileCost); + + InitialOffsetStore initialOffsetStore = InitialOffsetStore.getInstance(table, checkpointLocation); + this.initialOffset = getOrWriteInitialOffset(initialOffsetStore); + } + + @Override + public Offset latestOffset() { Review comment: **rationale**: streaming an **year old or a multi-year old iceberg table** - could potentially bring a frustrating experience. So, we will by default support `stream-start-mode` to be `incremental` and provide an option to specify `full`. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
