Spark SQL执行Join查询时遭遇scala.MatchError问题求助

在Amazon AWS环境运行Spark 3.0.1版本，执行带CTE的Join查询时触发scala.MatchError。单独查询table1和table2可正常返回结果，但执行Join操作报错，调换关联表顺序也无法解决。

使用的查询语句 #

query = '''
    WITH
      a1 AS (
         SELECT val_in_min, val_date
           FROM table1
      ),
      a2 AS (
         SELECT val_in_sec, val_date
           FROM table2
      )
    SELECT a.val_in_min, a.val_date, b.val_in_sec
      FROM a1 a
      JOIN a2 b
        ON a.val_date=b.val_date
'''
spark.sql(query).show()

样本数据 #

table1 #

table2 #

错误信息 #

An error was encountered:
An error occurred while calling o120.showString.
: scala.MatchError: list#77341 [] (of class org.apache.spark.sql.catalyst.expressions.ListQuery)
    at org.apache.spark.sql.catalyst.optimizer.SizeBasedJoinReorder$.hasFilterPredicate(SizeBasedJoinReorder.scala:285)
    at scala.collection.LinearSeqOptimized.exists(LinearSeqOptimized.scala:95)
    at scala.collection.LinearSeqOptimized.exists$(LinearSeqOptimized.scala:92)
    at scala.collection.immutable.List.exists(List.scala:89)
    at py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244)
    at py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357)
    at py4j.Gateway.invoke(Gateway.java:282)
    at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132)
    at py4j.commands.CallCommand.execute(CallCommand.java:79)
    at py4j.GatewayConnection.run(GatewayConnection.java:238)
    at java.lang.Thread.run(Thread.java:750)

Traceback (most recent call last):
  File "/mnt/yarn/usercache/appcache/application_1686133156822_0013/container_1686133156822_0013_01_000001/pyspark.zip/pyspark/sql/dataframe.py", line 441, in show
    print(self._jdf.showString(n, 20, vertical))
  File "/mnt/yarn/usercache/appcache/application_1686133156822_0013/container_1686133156822_0013_01_000001/py4j-0.10.9-src.zip/py4j/java_gateway.py", line 1305, in __call__
    answer, self.gateway_client, self.target_id, self.name)
  File "/mnt/yarn/usercache/appcache/application_1686133156822_0013/container_1686133156822_0013_01_000001/pyspark.zip/pyspark/sql/utils.py", line 128, in deco
    return f(*a, **kw)
  File "/mnt/yarn/usercache/appcache/application_1686133156822_0013/container_1686133156822_0013_01_000001/py4j-0.10.9-src.zip/py4j/protocol.py", line 328, in get_return_value
    format(target_id, ".", name), value)
py4j.protocol.Py4JJavaError: An error occurred while calling o120.showString.
: scala.MatchError: list#77341 [] (of class org.apache.spark.sql.catalyst.expressions.ListQuery)
    at org.apache.spark.sql.catalyst.optimizer.SizeBasedJoinReorder$.hasFilterPredicate(SizeBasedJoinReorder.scala:285)
    at org.apache.spark.sql.catalyst.optimizer.SizeBasedJoinReorder$.$anonfun$hasFilterPredicate$1(SizeBasedJoinReorder.scala:293)
    at org.apache.spark.sql.catalyst.optimizer.SizeBasedJoinReorder$.$anonfun$hasFilterPredicate$1$adapted(SizeBasedJoinReorder.scala:293)
    at scala.collection.LinearSeqOptimized.exists(LinearSeqOptimized.scala:95)
    at scala.collection.LinearSeqOptimized.exists$(LinearSeqOptimized.scala:92)
    at scala.collection.immutable.List.exists(List.scala:89)

解决建议 #

这是Spark 3.0.1版本中SizeBasedJoinReorder优化器的已知bug，可通过以下方式解决：

• 禁用该优化规则：在Spark配置中添加spark.sql.optimizer.excludedRules=org.apache.spark.sql.catalyst.optimizer.SizeBasedJoinReorder，禁用后查询可正常执行。
• 改写查询语句：去掉CTE结构，改用直接子查询的方式，例如：

  SELECT a.val_in_min, a.val_date, b.val_in_sec
  FROM (SELECT val_in_min, val_date FROM table1) a
  JOIN (SELECT val_in_sec, val_date FROM table2) b
  ON a.val_date = b.val_date
  

• 升级Spark版本：该bug在Spark 3.0.2及后续版本中已修复，升级到更高版本可彻底解决问题。

内容的提问来源于stack exchange，提问作者vvazza