无法通过 Hive Beeline 命令对 ORC 表的行数进行 select 计数

Question

I am using the following components - Hadoop 3.1.4, Hive 3.1.3 and Tez 0.9.2 And there is an ORC table from which I am trying to extract count of the rows in the table. select count(*) from ORC_TABLE and this throws the below set of exceptions

Error: Error while processing statement: FAILED: Execution Error, return code 2 from org.apache.hadoop.hive.ql.exec.tez.TezTask. Vertex failed, vertexName=Map 1, vertexId=vertex_1670915386694_0182_1_00, diagnostics=[Vertex vertex_1670915386694_0182_1_00 [Map 1] killed/failed due to:ROOT_INPUT_INIT_FAILURE, Vertex Input: jio_ar_consumer_events initializer failed, vertex=vertex_1670915386694_0182_1_00 [Map 1], java.lang.RuntimeException: ORC split generation failed with exception: java.lang.NoSuchMethodError: org.apache.hadoop.fs.FileStatus.compareTo(Lorg/apache/hadoop/fs/FileStatus;)I
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat.generateSplitsInfo(OrcInputFormat.java:1851)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat.getSplits(OrcInputFormat.java:1939)
    at org.apache.hadoop.hive.ql.io.HiveInputFormat.addSplitsForGroup(HiveInputFormat.java:519)
    at org.apache.hadoop.hive.ql.io.HiveInputFormat.getSplits(HiveInputFormat.java:765)
    at org.apache.hadoop.hive.ql.exec.tez.HiveSplitGenerator.initialize(HiveSplitGenerator.java:243)
    at org.apache.tez.dag.app.dag.RootInputInitializerManager$InputInitializerCallable$1.run(RootInputInitializerManager.java:278)
    at org.apache.tez.dag.app.dag.RootInputInitializerManager$InputInitializerCallable$1.run(RootInputInitializerManager.java:269)

Caused by: java.util.concurrent.ExecutionException: java.lang.NoSuchMethodError: org.apache.hadoop.fs.FileStatus.compareTo(Lorg/apache/hadoop/fs/FileStatus;)I
    at java.util.concurrent.FutureTask.report(FutureTask.java:122)
    at java.util.concurrent.FutureTask.get(FutureTask.java:192)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat.generateSplitsInfo(OrcInputFormat.java:1790)
    ... 17 more
Caused by: java.lang.NoSuchMethodError: org.apache.hadoop.fs.FileStatus.compareTo(Lorg/apache/hadoop/fs/FileStatus;)I
    at org.apache.hadoop.hive.ql.io.AcidUtils.lambda$getAcidState$0(AcidUtils.java:1117)
    at java.util.TimSort.countRunAndMakeAscending(TimSort.java:355)
    at java.util.TimSort.sort(TimSort.java:220)
    at java.util.Arrays.sort(Arrays.java:1512)
    at java.util.ArrayList.sort(ArrayList.java:1464)
    at java.util.Collections.sort(Collections.java:177)
    at org.apache.hadoop.hive.ql.io.AcidUtils.getAcidState(AcidUtils.java:1115)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat$FileGenerator.callInternal(OrcInputFormat.java:1207)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat$FileGenerator.access$1500(OrcInputFormat.java:1142)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat$FileGenerator$1.run(OrcInputFormat.java:1179)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat$FileGenerator$1.run(OrcInputFormat.java:1176)
    at java.security.AccessController.doPrivileged(Native Method)
    at javax.security.auth.Subject.doAs(Subject.java:422)
    at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1657)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat$FileGenerator.call(OrcInputFormat.java:1176)
    at org.apache.hadoop.hive.ql.io.orc.OrcInputFormat$FileGenerator.call(OrcInputFormat.java:1142)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    ... 3 more

There is another article where the same problem has been described ORC Split Generation issue with Hive Table but there isnt any solution as such yet. I also tried running CONCATENATE function on top of ORC Table but that didn't help either.

What works though is, if I run select * from ORC_TABLE with or without LIMIT, it seems to extract the records. I reckon issue must only be with aggregate functions or may be I don't get the issue yet.

I am also using Spark 3.3.1 and I can extract the same count through Spark Context Spark Sql utility and able to fetch the rows as well. No issues with Spark in that front.

Adding on to it, When I change the execution engine to MR, then this works. Fails only when I run this on Tez Engine.

Any leads to resolve this issue is much appreciated.

Answer 1

The issue was resolved by the below steps based my previous analysis:

This class org.apache.hadoop.fs.FileStatus comes as a part of hadoop common jar file.

We were using Hadoop 3.1.4 & Tez 0.9.2

Tez 0.9.2 contains a tez.tar.gz that needs to be placed onto HDFS location. This tez.tar.gz contained hadoop-common-2.7.2.jar (This does not have the method compareTo that is thrown as an exception as shown in the error )

Solution:

We extracted the tez.tar.gz and replaced all hadoop 2.7.2 related jars with hadoop 3.1.4 jars. Do this if you dont want to reconfigure again with new tez version. Otherwise you could follow solution 2 as mentioned.

Recreated the tar and placed it across all dependent locations including HDFS as well. For us it was in /user/tez/share/tez.tar.gz location. It changes accordingly.

This error disappeared after I followed the steps and now I am able to do count of records on any table.

Solution 2: Other solution that you could easily do is, use 0.10.x Tez version that contains libraries for hadoop 3.x version. Rather than 0.9.2 Tez version which is compatible with hadoop 2.7.x version.