HIVE-24225: Revert HIVE-23158 Optimize S3A recordReader policy for Random IO formats

llap-server/src/java/org/apache/hadoop/hive/llap/io/api/impl/LlapInputFormat.java

@@ -20,8 +20,6 @@
 package org.apache.hadoop.hive.llap.io.api.impl;
 
 import org.apache.hadoop.fs.FileSystem;
-import org.apache.hadoop.fs.s3a.S3AFileSystem;
-import org.apache.hadoop.fs.s3a.S3AInputPolicy;
 import org.apache.hadoop.hive.ql.exec.vector.VectorizedSupport;
 import org.apache.hadoop.hive.ql.io.BatchToRowInputFormat;
 import org.apache.hadoop.hive.conf.HiveConf;
@@ -65,9 +63,6 @@
 import org.apache.hadoop.mapred.Reporter;
 import org.apache.hive.common.util.HiveStringUtils;
 
-import static org.apache.hadoop.hive.common.FileUtils.isS3a;
-import static org.apache.hadoop.hive.ql.io.HiveInputFormat.isRandomAccessInputFormat;
-
 public class LlapInputFormat implements InputFormat<NullWritable, VectorizedRowBatch>,
     VectorizedInputFormatInterface, SelfDescribingInputFormatInterface,
     AvoidSplitCombination {
@@ -107,10 +102,6 @@ public RecordReader<NullWritable, VectorizedRowBatch> getRecordReader(
     FileSplit fileSplit = (FileSplit) split;
     reporter.setStatus(fileSplit.toString());
     FileSystem splitFileSystem = fileSplit.getPath().getFileSystem(job);
-    if (isS3a(splitFileSystem) && isRandomAccessInputFormat(sourceInputFormat)) {
-      LlapIoImpl.LOG.debug("Changing S3A input policy to RANDOM");
-      ((S3AFileSystem) splitFileSystem).setInputPolicy(S3AInputPolicy.Random);
-    }
     try {
       // At this entry point, we are going to assume that these are logical table columns.
       // Perhaps we should go thru the code and clean this up to be more explicit; for now, we

ql/pom.xml

@@ -321,11 +321,6 @@
       <artifactId>hadoop-yarn-client</artifactId>
       <optional>true</optional>
     </dependency>
-    <dependency>
-      <groupId>org.apache.hadoop</groupId>
-      <artifactId>hadoop-aws</artifactId>
-      <version>${hadoop.version}</version>
-    </dependency>
     <dependency>
       <groupId>org.apache.orc</groupId>
       <artifactId>orc-tools</artifactId>

ql/src/java/org/apache/hadoop/hive/ql/io/HiveInputFormat.java

@@ -24,8 +24,6 @@
 import org.apache.hadoop.fs.FileStatus;
 import org.apache.hadoop.fs.FileSystem;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.fs.s3a.S3AFileSystem;
-import org.apache.hadoop.fs.s3a.S3AInputPolicy;
 import org.apache.hadoop.hive.common.FileUtils;
 import org.apache.hadoop.hive.common.StringInternUtils;
 import org.apache.hadoop.hive.common.ValidTxnWriteIdList;
@@ -40,8 +38,6 @@
 import org.apache.hadoop.hive.ql.exec.TableScanOperator;
 import org.apache.hadoop.hive.ql.exec.Utilities;
 import org.apache.hadoop.hive.ql.exec.vector.VectorizedRowBatch;
-import org.apache.hadoop.hive.ql.io.orc.OrcInputFormat;
-import org.apache.hadoop.hive.ql.io.parquet.VectorizedParquetInputFormat;
 import org.apache.hadoop.hive.ql.log.PerfLogger;
 import org.apache.hadoop.hive.ql.metadata.HiveException;
 import org.apache.hadoop.hive.ql.metadata.HiveStoragePredicateHandler;
@@ -93,7 +89,6 @@
 import java.util.concurrent.Future;
 
 import static java.lang.Integer.min;
-import static org.apache.hadoop.hive.common.FileUtils.isS3a;
 
 /**
  * HiveInputFormat is a parameterized InputFormat which looks at the path name
@@ -384,19 +379,6 @@ public static InputFormat<WritableComparable, Writable> getInputFormatFromCache(
     return instance;
   }
 
-  /**
-   * Returns true if the inputFormat performs random seek+read
-   * @param inputFormat
-   * @return
-   */
-  public static boolean isRandomAccessInputFormat(InputFormat inputFormat) {
-    if (inputFormat instanceof OrcInputFormat ||
-        inputFormat instanceof VectorizedParquetInputFormat) {
-      return true;
-    }
-    return false;
-  }
-
   @Override
   public RecordReader getRecordReader(InputSplit split, JobConf job,
       Reporter reporter) throws IOException {
@@ -449,13 +431,6 @@ public RecordReader getRecordReader(InputSplit split, JobConf job,
       innerReader = HiveIOExceptionHandlerUtil
           .handleRecordReaderCreationException(e, job);
     }
-
-    FileSystem splitFileSystem = splitPath.getFileSystem(job);
-    if (isS3a(splitFileSystem) && isRandomAccessInputFormat(inputFormat)) {
-      LOG.debug("Changing S3A input policy to RANDOM");
-      ((S3AFileSystem) splitFileSystem).setInputPolicy(S3AInputPolicy.Random);
-    }
-
     HiveRecordReader<K,V> rr = new HiveRecordReader(innerReader, job);
     rr.initIOContext(hsplit, job, inputFormatClass, innerReader);
     return rr;