How does hive decide on the insert query plan

[#1]

We're working with Hive 1.3.1, and running an INSERT statement to upload data into Hive from an external table.

I noticed that the execution plan has changed on the same table from yesterday compared to today.

Yesterday the plan resulted in a M/R job with 341 mappers and 359 reducers, while today the plan resulted in a M/R job with only mappers and no reducers

This is the query:

explain insert OVERWRITE table managed_table PARTITION(col1) select [columns] from external_table

How does hive decide how to execute the query?

How does it translate the insert select into map reduce?

What would cause a plan to change?

This is the first plan (omitting columns lists because the table has over 300 columns)

<code>STAGE DEPENDENCIES:

  Stage-1 is a root stage

  Stage-0 depends on stages: Stage-1

  Stage-2 depends on stages: Stage-0

STAGE PLANS:

  Stage: Stage-1

    Map Reduce

      Map Operator Tree:

          TableScan
            alias: externalevents
            Statistics: Num rows: 479391 Data size: 91824480256 Basic stats: COMPLETE Column stats: NONE
            Select Operator  [columns] outputColumnNames:  [columns]  Statistics: Num rows: 479391 Data size: 91824480256 Basic stats: COMPLETE Column stats: NONE
              Reduce Output Operator
                key expressions: _col394 (type: bigint)
                sort order: +
                Map-reduce partition columns: _col394 (type: bigint)
                Statistics: Num rows: 479391 Data size: 91824480256 Basic stats: COMPLETE Column stats: NONE
                value expressions: [columns]  Reduce Operator Tree:
        Select Operator
          expressions: [columns]  outputColumnNames: [columns] Statistics: Num rows: 479391 Data size: 91824480256 Basic stats: COMPLETE Column stats: NONE
          File Output Operator
            compressed: false
            Statistics: Num rows: 479391 Data size: 91824480256 Basic stats: COMPLETE Column stats: NONE
            table:
                input format: org.apache.hadoop.hive.ql.io.orc.OrcInputFormat
                output format: org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat
                serde: org.apache.hadoop.hive.ql.io.orc.OrcSerde
                name: ceazip.events_test_hive

  Stage: Stage-0
    Move Operator
      tables:
          partition:
            evtf_first_date_id
          replace: true
          table:
              input format: org.apache.hadoop.hive.ql.io.orc.OrcInputFormat
              output format: org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat
              serde: org.apache.hadoop.hive.ql.io.orc.OrcSerde
              name: ceazip.events_test_hive

  Stage: Stage-2
    Stats-Aggr Operator

<code>And the 2nd plan:

<code>STAGE PLANS:
  Stage: Stage-1
    Map Reduce
      Map Operator Tree:
          TableScan
            alias: events_2017_02_21_11_13_18
            Statistics: Num rows: 468680 Data size: 89772957696 Basic stats: COMPLETE Column stats: NONE
            Select Operator
              expressions: [columns] 
outputColumnNames: [columns] 
 Statistics: Num rows: 468680 Data size: 89772957696 Basic stats: COMPLETE Column stats: NONE
              File Output Operator
                compressed: false
                Statistics: Num rows: 468680 Data size: 89772957696 Basic stats: COMPLETE Column stats: NONE
                table:
                    input format: org.apache.hadoop.hive.ql.io.orc.OrcInputFormat
                    output format: org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat
                    serde: org.apache.hadoop.hive.ql.io.orc.OrcSerde
                    name: default.events_test1
  Stage: Stage-7
    Conditional Operator
  Stage: Stage-4
    Move Operator
      files:
          hdfs directory: true
          destination: hdfs://...../apps/hive/warehouse/events_test1/.hive-staging_hive_2017-03-01_07-35-18_776_4958999242494325333-1/-ext-10000
  Stage: Stage-0
    Move Operator
      tables:
          partition:
            evtf_first_date_id
          replace: true
          table:
              input format: org.apache.hadoop.hive.ql.io.orc.OrcInputFormat
              output format: org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat
              serde: org.apache.hadoop.hive.ql.io.orc.OrcSerde
              name: default.events_test1
  Stage: Stage-2
    Stats-Aggr Operator
  Stage: Stage-3
    Merge File Operator
      Map Operator Tree:
          ORC File Merge Operator
      merge level: stripe
      input format: org.apache.hadoop.hive.ql.io.orc.OrcInputFormat
  Stage: Stage-5
    Merge File Operator
      Map Operator Tree:
          ORC File Merge Operator
      merge level: stripe
      input format: org.apache.hadoop.hive.ql.io.orc.OrcInputFormat
  Stage: Stage-6
    Move Operator
      files:
          hdfs directory: true
          destination: hdfs://isr-r0-aps-nam-1.lab.il.nice.com:8020/apps/hive/warehouse/events_test1/.hive-staging_hive_2017-03-01_07-35-18_776_4958999242494325333-1/-ext-10000

Thanks,
Lior

[#2 · Accepted solution]

@Lior Hadaya

CBO (cost based optimizer) and statistics collected on your tables.

You may have the settings mentioned here set to true: https://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.2/bk_performance_tuning/content/hive_perf_bes...

As such, the behavior can change over time. You could also force stats on a specific table or even column.