[GitHub] [hudi] alexeykudinkin commented on a diff in pull request #7818: [HUDI-5678] Fix `deduceShuffleParallelism` in row-writing Bulk Insert helper

["alexeykudinkin (via GitHub)" <gi...@apache.org>]

alexeykudinkin commented on code in PR #7818:
URL: https://github.com/apache/hudi/pull/7818#discussion_r1094800871


##########
hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestInsertTable.scala:
##########
@@ -1159,4 +1161,68 @@ class TestInsertTable extends HoodieSparkSqlTestBase {
       )
     }
   }
+
+  /**
+   * This test is to make sure that bulk insert doesn't create a bunch of tiny files if
+   * hoodie.bulkinsert.user.defined.partitioner.sort.columns doesn't start with the partition columns
+   *
+   * NOTE: Additionally, this test serves as a smoke test making sure that all of the bulk-insert
+   *       modes work
+   */
+  test(s"Test Bulk Insert with all sort-modes") {
+    withTempDir { basePath =>
+      BulkInsertSortMode.values().foreach { sortMode =>
+        val tableName = generateTableName
+        // Remove these with [HUDI-5419]
+        spark.sessionState.conf.unsetConf("hoodie.datasource.write.operation")
+        spark.sessionState.conf.unsetConf("hoodie.datasource.write.insert.drop.duplicates")
+        spark.sessionState.conf.unsetConf("hoodie.merge.allow.duplicate.on.inserts")
+        spark.sessionState.conf.unsetConf("hoodie.datasource.write.keygenerator.consistent.logical.timestamp.enabled")
+        // Default parallelism is 200 which means in global sort, each record will end up in a different spark partition so
+        // 9 files would be created. Setting parallelism to 3 so that each spark partition will contain a hudi partition.
+        val parallelism = if (sortMode.name.equals(BulkInsertSortMode.GLOBAL_SORT.name())) {
+          "hoodie.bulkinsert.shuffle.parallelism = 3,"
+        } else {
+          ""
+        }
+        spark.sql(
+          s"""
+             |create table $tableName (
+             |  id int,
+             |  name string,
+             |  price double,
+             |  dt string
+             |) using hudi
+             | tblproperties (
+             |  primaryKey = 'id',
+             |  preCombineField = 'name',
+             |  type = 'cow',
+             |  $parallelism
+             |  hoodie.bulkinsert.sort.mode = '${sortMode.name}'
+             | )
+             | partitioned by (dt)
+             | location '${basePath.getCanonicalPath}/$tableName'
+                """.stripMargin)
+
+        spark.sql("set hoodie.sql.bulk.insert.enable = true")
+        spark.sql("set hoodie.sql.insert.mode = non-strict")
+
+        spark.sql(
+          s"""insert into $tableName  values
+             |(5, 'a', 35, '2021-05-21'),
+             |(1, 'a', 31, '2021-01-21'),
+             |(3, 'a', 33, '2021-03-21'),
+             |(4, 'b', 16, '2021-05-21'),
+             |(2, 'b', 18, '2021-01-21'),
+             |(6, 'b', 17, '2021-03-21'),
+             |(8, 'a', 21, '2021-05-21'),
+             |(9, 'a', 22, '2021-01-21'),
+             |(7, 'a', 23, '2021-03-21')
+             |""".stripMargin)
+
+        // TODO re-enable
+        //assertResult(3)(spark.sql(s"select distinct _hoodie_file_name from $tableName").count())

Review Comment:
   This test is failing currently
   
   @jonvex will follow-up to enable it



-- 
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.

To unsubscribe, e-mail: commits-unsubscribe@hudi.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org