stevenzwu commented on a change in pull request #1185: URL: https://github.com/apache/iceberg/pull/1185#discussion_r479457104
########## File path: flink/src/main/java/org/apache/iceberg/flink/sink/IcebergFilesCommitter.java ########## @@ -0,0 +1,229 @@ +/* + * 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.flink.sink; + +import java.util.Comparator; +import java.util.List; +import java.util.Map; +import java.util.NavigableMap; +import java.util.SortedMap; +import org.apache.flink.api.common.state.ListState; +import org.apache.flink.api.common.state.ListStateDescriptor; +import org.apache.flink.api.common.typeinfo.BasicTypeInfo; +import org.apache.flink.api.common.typeinfo.TypeInformation; +import org.apache.flink.api.java.typeutils.ListTypeInfo; +import org.apache.flink.runtime.state.StateInitializationContext; +import org.apache.flink.runtime.state.StateSnapshotContext; +import org.apache.flink.streaming.api.operators.AbstractStreamOperator; +import org.apache.flink.streaming.api.operators.BoundedOneInput; +import org.apache.flink.streaming.api.operators.OneInputStreamOperator; +import org.apache.flink.streaming.runtime.streamrecord.StreamRecord; +import org.apache.flink.table.runtime.typeutils.SortedMapTypeInfo; +import org.apache.hadoop.conf.Configuration; +import org.apache.iceberg.AppendFiles; +import org.apache.iceberg.DataFile; +import org.apache.iceberg.Snapshot; +import org.apache.iceberg.Table; +import org.apache.iceberg.flink.TableLoader; +import org.apache.iceberg.hadoop.SerializableConfiguration; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.relocated.com.google.common.collect.ImmutableList; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.relocated.com.google.common.collect.Maps; +import org.apache.iceberg.types.Comparators; +import org.apache.iceberg.types.Types; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +class IcebergFilesCommitter extends AbstractStreamOperator<Void> + implements OneInputStreamOperator<DataFile, Void>, BoundedOneInput { + + private static final long serialVersionUID = 1L; + private static final long INITIAL_CHECKPOINT_ID = -1L; + + private static final Logger LOG = LoggerFactory.getLogger(IcebergFilesCommitter.class); + private static final String FLINK_JOB_ID = "flink.job-id"; + + // The max checkpoint id we've committed to iceberg table. As the flink's checkpoint is always increasing, so we could + // correctly commit all the data files whose checkpoint id is greater than the max committed one to iceberg table, for + // avoiding committing the same data files twice. This id will be attached to iceberg's meta when committing the + // iceberg transaction. + private static final String MAX_COMMITTED_CHECKPOINT_ID = "flink.max-committed-checkpoint-id"; + + // TableLoader to load iceberg table lazily. + private final TableLoader tableLoader; + private final SerializableConfiguration hadoopConf; + + // A sorted map to maintain the completed data files for each pending checkpointId (which have not been committed + // to iceberg table). We need a sorted map here because there's possible that few checkpoints snapshot failed, for + // example: the 1st checkpoint have 2 data files <1, <file0, file1>>, the 2st checkpoint have 1 data files + // <2, <file3>>. Snapshot for checkpoint#1 interrupted because of network/disk failure etc, while we don't expect + // any data loss in iceberg table. So we keep the finished files <1, <file0, file1>> in memory and retry to commit + // iceberg table when the next checkpoint happen. + private final NavigableMap<Long, List<DataFile>> dataFilesPerCheckpoint = Maps.newTreeMap(); Review comment: commit failure certain isn't common. Agree that it shouldn't be a blocker for the initial version. Just to provide more context on why it is important to us. Our data warehouses (metastore) only lives in us-east-1 region, while Flink streaming jobs can run in 3 regions (us-east-1, us-west-2, and eu-west-1). As Ryan mentioned, we are more concerned about extended outages (like a day) for whatever reason (us-east-1 outage, cross-region network issue, metastore service outage). For high-parallelism or event time partitioned tables, there could be thousands or tens of thousands of files per checkpoint interval. This manifest approach allows the Flink jobs to handle those extended outages better. Flink operator list state can't handle large state well. I vaguely remember 1 or 2 GBs. And it can get pretty slow when the list is large. ---------------------------------------------------------------- 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: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
