[jira] [Created] (HUDI-4113) Cannot parse schema when use spark delete sql

["KnightChess (Jira)" <ji...@apache.org>]

KnightChess created HUDI-4113:
---------------------------------

             Summary: Cannot parse <null> schema when use spark delete sql
                 Key: HUDI-4113
                 URL: https://issues.apache.org/jira/browse/HUDI-4113
             Project: Apache Hudi
          Issue Type: Bug
          Components: spark-sql
            Reporter: KnightChess
            Assignee: KnightChess


cause `Cannot parse <null> schema` when use delete sql after alter drop partitions

#### Devices：

```sql
CREATE  TABLE delete_error_test(
     id int,
     age int,
     name string,
     dt string)
using hudi
options (
     type = 'cow',
     primaryKey = 'id',
     hoodie.table.precombine.field = 'age',
     hoodie.datasource.query.type = 'snapshot',
     hoodie.metadata.enable = 'true',
     hoodie.metadata.compact.max.delta.commits = '1',
     hoodie.metadata.keep.min.commits = '2',
     hoodie.metadata.keep.max.commits = '3',
     hoodie.cleaner.policy = 'KEEP_LATEST_COMMITS',
     hoodie.cleaner.commits.retained = '3',
     hoodie.datasource.meta.sync.enable = 'true'
)
partitioned by (dt)";
 
 
insert overwrite delete_error_test select 1, 18, 'wlq', '20220101';
insert overwrite delete_error_test select 2, 19, 'wlq', '20220102';
alter table delete_error_test drop partition (dt = '20220101');
 
# after drop other partitions, this sql will trigger error
delete from delete_error_test where dt = '20220102';
```

#### stack:

```shell
Caused by: org.apache.hudi.exception.HoodieUpsertException: Error upserting bucketType UPDATE for partition :0
    at org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.handleUpsertPartition(BaseSparkCommitActionExecutor.java:329)
    at org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.lambda$mapPartitionsAsRDD$a3ab3c4$1(BaseSparkCommitActionExecutor.java:244)
    at org.apache.spark.api.java.JavaRDDLike.$anonfun$mapPartitionsWithIndex$1(JavaRDDLike.scala:102)
    at org.apache.spark.api.java.JavaRDDLike.$anonfun$mapPartitionsWithIndex$1$adapted(JavaRDDLike.scala:102)
    at org.apache.spark.rdd.RDD.$anonfun$mapPartitionsWithIndex$2(RDD.scala:915)
    at org.apache.spark.rdd.RDD.$anonfun$mapPartitionsWithIndex$2$adapted(RDD.scala:915)
    at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:52)
    at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:373)
    at org.apache.spark.rdd.RDD.iterator(RDD.scala:337)
    at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:52)
    at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:373)
    at org.apache.spark.rdd.RDD.$anonfun$getOrCompute$1(RDD.scala:386)
    at org.apache.spark.storage.BlockManager.$anonfun$doPutIterator$1(BlockManager.scala:1498)
    at org.apache.spark.storage.BlockManager.org$apache$spark$storage$BlockManager$doPut(BlockManager.scala:1408)
    at org.apache.spark.storage.BlockManager.doPutIterator(BlockManager.scala:1472)
    at org.apache.spark.storage.BlockManager.getOrElseUpdate(BlockManager.scala:1295)
    at org.apache.spark.rdd.RDD.getOrCompute(RDD.scala:384)
    at org.apache.spark.rdd.RDD.iterator(RDD.scala:335)
    at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:52)
    at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:373)
    at org.apache.spark.rdd.RDD.iterator(RDD.scala:337)
    at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:90)
    at org.apache.spark.scheduler.Task.run(Task.scala:131)
    at org.apache.spark.executor.Executor$TaskRunner.$anonfun$run$3(Executor.scala:506)
    at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1462)
    at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:509)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.avro.SchemaParseException: Cannot parse <null> schema
    at org.apache.avro.Schema.parse(Schema.java:1633)
    at org.apache.avro.Schema$Parser.parse(Schema.java:1430)
    at org.apache.avro.Schema$Parser.parse(Schema.java:1418)
    at org.apache.hudi.io.HoodieWriteHandle.getSpecifiedTableSchema(HoodieWriteHandle.java:137)
    at org.apache.hudi.io.HoodieWriteHandle.lambda$new$0(HoodieWriteHandle.java:119)
    at org.apache.hudi.common.util.Option.orElseGet(Option.java:142)
    at org.apache.hudi.io.HoodieWriteHandle.<init>(HoodieWriteHandle.java:119)
    at org.apache.hudi.io.HoodieWriteHandle.<init>(HoodieWriteHandle.java:109)
    at org.apache.hudi.io.HoodieMergeHandle.<init>(HoodieMergeHandle.java:123)
    at org.apache.hudi.io.HoodieMergeHandle.<init>(HoodieMergeHandle.java:116)
    at org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.getUpdateHandle(BaseSparkCommitActionExecutor.java:377)
    at org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.handleUpdate(BaseSparkCommitActionExecutor.java:348)
    at org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.handleUpsertPartition(BaseSparkCommitActionExecutor.java:322)
    ... 28 more
```



--
This message was sent by Atlassian Jira
(v8.20.7#820007)