[
https://issues.apache.org/jira/browse/BEAM-7695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17547988#comment-17547988
]
Danny McCormick commented on BEAM-7695:
---------------------------------------
This issue has been migrated to https://github.com/apache/beam/issues/19644
> Read TFRecord Files from hdfs will meet exception if file size is large
> -----------------------------------------------------------------------
>
> Key: BEAM-7695
> URL: https://issues.apache.org/jira/browse/BEAM-7695
> Project: Beam
> Issue Type: Bug
> Components: io-java-tfrecord, runner-flink
> Affects Versions: 2.2.0, 2.3.0, 2.4.0, 2.5.0, 2.6.0, 2.7.0, 2.8.0, 2.9.0,
> 2.10.0, 2.11.0, 2.12.0, 2.13.0
> Reporter: LI HAO
> Priority: P3
>
> I read the TFRecord files which is in HDFS will meet error.
> * The single TFRecord file is larger than 3GB.
> * The total size larger than 1TB.
> * Using Beam 2.13.0 + Flinkrunner 2.13.0 + Java 1.8, I also test
> 2.11.0/2.12.0 with same problem
> The dependency jar (in build.gradle):
> {code:java}
> dependencies
> { // This dependency is found on compile classpath of this component and
> consumers. //implementation 'com.google.guava:guava:27.0.1-jre' compile
> 'org.apache.beam:beam-sdks-java-core:2.13.0' compile
> 'org.apache.beam:beam-vendor-guava-20_0:0.1' compile
> 'org.tensorflow:tensorflow-hadoop:1.13.1' compile
> 'org.apache.beam:beam-runners-direct-java:2.13.0' //implementation
> "org.apache.beam:beam-sdks-java-core:2.13.0" compile
> "org.apache.beam:beam-runners-flink_2.11:2.13.0" compile
> "org.apache.beam:beam-sdks-java-io-hadoop-file-system:2.13.0" compile
> "org.apache.hadoop:hadoop-common:2.7.3" compile
> "org.apache.hadoop:hadoop-client:2.7.3" compile
> "org.apache.hadoop:hadoop-mapreduce-client-core:2.7.3" compile
> "org.tensorflow:proto:1.13.1" compile
> "org.apache.beam:beam-sdks-java-extensions-sketching:2.13.0" // Use JUnit
> test framework testImplementation 'junit:junit:4.12' }{code}
> The error msg:
>
> {code:java}
> ------------------------------------------------------------
> The program finished with the following exception:
> org.apache.flink.client.program.ProgramInvocationException: The main method
> caused an error.
> at
> org.apache.flink.client.program.PackagedProgram.callMainMethod(PackagedProgram.java:546)
> at
> org.apache.flink.client.program.PackagedProgram.invokeInteractiveModeForExecution(PackagedProgram.java:421)
> at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:427)
> at
> org.apache.flink.client.cli.CliFrontend.executeProgram(CliFrontend.java:813)
> at org.apache.flink.client.cli.CliFrontend.runProgram(CliFrontend.java:287)
> at org.apache.flink.client.cli.CliFrontend.run(CliFrontend.java:213)
> at
> org.apache.flink.client.cli.CliFrontend.parseParameters(CliFrontend.java:1050)
> at
> org.apache.flink.client.cli.CliFrontend.lambda$main$11(CliFrontend.java:1126)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:422)
> at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1692)
> at
> org.apache.flink.runtime.security.HadoopSecurityContext.runSecured(HadoopSecurityContext.java:41)
> at org.apache.flink.client.cli.CliFrontend.main(CliFrontend.java:1126)
> Caused by: java.lang.RuntimeException: Pipeline execution failed
> at org.apache.beam.runners.flink.FlinkRunner.run(FlinkRunner.java:116)
> at org.apache.beam.sdk.Pipeline.run(Pipeline.java:313)
> at org.apache.beam.sdk.Pipeline.run(Pipeline.java:299)
> at avazu.data.transform.App.testTfrecordQIYU(App.java:572)
> at avazu.data.transform.App.main(App.java:744)
> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
> at
> sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:498)
> at
> org.apache.flink.client.program.PackagedProgram.callMainMethod(PackagedProgram.java:529)
> ... 12 more
> Caused by: org.apache.flink.client.program.ProgramInvocationException: Job
> failed. (JobID: fe4ce5375efbbb55e56967e4c7a975b2)
> at
> org.apache.flink.client.program.rest.RestClusterClient.submitJob(RestClusterClient.java:268)
> at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:487)
> at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:475)
> at
> org.apache.flink.client.program.ContextEnvironment.execute(ContextEnvironment.java:62)
> at
> org.apache.beam.runners.flink.FlinkPipelineExecutionEnvironment.executePipeline(FlinkPipelineExecutionEnvironment.java:139)
> at org.apache.beam.runners.flink.FlinkRunner.run(FlinkRunner.java:113)
> ... 21 more
> Caused by: org.apache.flink.runtime.client.JobExecutionException: Job
> execution failed.
> at
> org.apache.flink.runtime.jobmaster.JobResult.toJobExecutionResult(JobResult.java:146)
> at
> org.apache.flink.client.program.rest.RestClusterClient.submitJob(RestClusterClient.java:265)
> ... 26 more
> Caused by: java.io.IOException: Mismatch of length mask when reading a
> record. Expected 808268081 but received 1769712859.
> at org.apache.beam.sdk.io.TFRecordIO$TFRecordCodec.read(TFRecordIO.java:651)
> at
> org.apache.beam.sdk.io.TFRecordIO$TFRecordSource$TFRecordReader.readNextRecord(TFRecordIO.java:530)
> at
> org.apache.beam.sdk.io.CompressedSource$CompressedReader.readNextRecord(CompressedSource.java:431)
> at
> org.apache.beam.sdk.io.FileBasedSource$FileBasedReader.advanceImpl(FileBasedSource.java:484)
> at
> org.apache.beam.sdk.io.FileBasedSource$FileBasedReader.startImpl(FileBasedSource.java:479)
> at
> org.apache.beam.sdk.io.OffsetBasedSource$OffsetBasedReader.start(OffsetBasedSource.java:249)
> at
> org.apache.beam.runners.flink.metrics.ReaderInvocationUtil.invokeStart(ReaderInvocationUtil.java:51)
> at
> org.apache.beam.runners.flink.translation.wrappers.SourceInputFormat.open(SourceInputFormat.java:75)
> at
> org.apache.beam.runners.flink.translation.wrappers.SourceInputFormat.open(SourceInputFormat.java:42)
> at
> org.apache.flink.runtime.operators.DataSourceTask.invoke(DataSourceTask.java:170)
> at org.apache.flink.runtime.taskmanager.Task.run(Task.java:704)
> at java.lang.Thread.run(Thread.java:745)
> {code}
> How to fix?
> * I have already fix the bug with the following code.
> * I will refine the following code and commit the codes.
>
> {code:java}
> diff --git
> a/sdks/java/core/src/main/java/org/apache/beam/sdk/io/TFRecordIO.java
> b/sdks/java/core/src/main/java/org/apache/beam/sdk/io/TFRecordIO.java
> index 96a753a..484a7cb 100644
> --- a/sdks/java/core/src/main/java/org/apache/beam/sdk/io/TFRecordIO.java
> +++ b/sdks/java/core/src/main/java/org/apache/beam/sdk/io/TFRecordIO.java
> @@ -631,8 +631,16 @@ public class TFRecordIO {
> int headerBytes = inChannel.read(header);
> if (headerBytes <= 0) {
> return null;
> + } else if (headerBytes != HEADER_LEN) {
> + while (header.hasRemaining() && inChannel.read(header) >= 0) {}
> + if (header.hasRemaining()) {
> + throw new IOException(String.format(
> + "EOF while reading record of length %d. Read only %d bytes. Input might be
> truncated. Not a valid TFRecord. Fewer than 12 bytes.",
> + HEADER_LEN, header.position()));
> + }
> + } else {
> +
> }
> - checkState(headerBytes == HEADER_LEN, "Not a valid TFRecord. Fewer than 12
> bytes.");
> header.rewind();
> long length = header.getLong();
> @@ -655,7 +663,12 @@ public class TFRecordIO {
> }
> footer.clear();
> - inChannel.read(footer);
> + while (footer.hasRemaining() && inChannel.read(footer) >= 0) {}
> + if (footer.hasRemaining()) {
> + throw new IOException(String.format(
> + "EOF while reading record of length %d. Read only %d bytes. Input might be
> truncated. Footer error.",
> + FOOTER_LEN, footer.position()));
> + }
> footer.rewind();
> int maskedCrc32OfData = footer.getInt();
> {code}
>
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
