wangzhongz commented on issue #4184:
URL: https://github.com/apache/hudi/issues/4184#issuecomment-985179200
hoodie configs:
override def getHudiConfig(): HudiConfigBuilder = {
super.getHudiConfig()
.properties(Map[String, String](
HoodieTableConfig.TYPE.key() -> MOR_TABLE_TYPE_OPT_VAL
,HoodieTableConfig.RECORDKEY_FIELDS.key() -> "start,end"
,HoodieTableConfig.PARTITION_FIELDS.key() -> "relation"
,HoodieTableConfig.PRECOMBINE_FIELD.key() -> "lastUTS"
,HoodieTableConfig.KEY_GENERATOR_CLASS_NAME.key() ->
classOf[ComplexKeyGenerator].getName
,TABLE_TYPE.key() -> MOR_TABLE_TYPE_OPT_VAL
,OPERATION.key() -> UPSERT_OPERATION_OPT_VAL
,RECORDKEY_FIELD.key() -> "start,end"
,PARTITIONPATH_FIELD.key() -> "relation"
,HIVE_STYLE_PARTITIONING.key() -> "true"
,RECONCILE_SCHEMA.key() -> "true"
,HoodieWriteConfig.PRECOMBINE_FIELD_NAME.key() -> "lastUTS"
,HoodieWriteConfig.BULKINSERT_PARALLELISM_VALUE.key() -> "200"
,HoodieWriteConfig.INSERT_PARALLELISM_VALUE.key() -> "200"
,HoodieWriteConfig.UPSERT_PARALLELISM_VALUE.key() -> "200"
,HoodieWriteConfig.BULK_INSERT_SORT_MODE.key() ->
BulkInsertSortMode.PARTITION_SORT.toString
,HoodieWriteConfig.WRITE_PAYLOAD_CLASS_NAME.key() ->
classOf[GraphHoodieRecordPayload].getName
,HoodieWriteConfig.KEYGENERATOR_TYPE.key() -> "COMPLEX"
,HoodieWriteConfig.KEYGENERATOR_CLASS_NAME.key() ->
classOf[ComplexKeyGenerator].getName
,HoodieIndexConfig.INDEX_TYPE.key() -> "BLOOM"
,HoodieIndexConfig.BLOOM_FILTER_NUM_ENTRIES_VALUE.key() ->
"200000"
,HoodiePayloadConfig.ORDERING_FIELD.key() -> "lastUTS"
,HoodiePayloadConfig.EVENT_TIME_FIELD.key() -> "lastUTS"
,HoodieStorageConfig.PARQUET_MAX_FILE_SIZE.key() ->
String.valueOf(128 * 1024 * 1024)
,HoodieStorageConfig.PARQUET_BLOCK_SIZE.key() ->
String.valueOf(128 * 1024 * 1024)
,HoodieStorageConfig.LOGFILE_MAX_SIZE.key() -> String.valueOf(1024
* 1024 * 1024)
,HoodieStorageConfig.LOGFILE_DATA_BLOCK_MAX_SIZE.key() ->
String.valueOf(256 * 1024 * 1024)
,HoodieCompactionConfig.PAYLOAD_CLASS_NAME.key() ->
classOf[GraphHoodieRecordPayload].getName
,HoodieCompactionConfig.ASYNC_CLEAN.key() -> "true"
,HoodieCompactionConfig.CLEANER_COMMITS_RETAINED.key() -> "3"
,HoodieCompactionConfig.INLINE_COMPACT_NUM_DELTA_COMMITS.key() ->
"1"
,HoodieCompactionConfig.INLINE_COMPACT.key() -> "true"
,HoodieCompactionConfig.CLEANER_PARALLELISM_VALUE.key() -> "200"
))
}
df.write.format("org.apache.hudi").
options(c.getHudiConfig().build()).
option(OPERATION.key(), "upsert").
option(HoodieWriteConfig.BULKINSERT_PARALLELISM_VALUE.key()
,argument.parallelism.toInt ).
option(HoodieWriteConfig.INSERT_PARALLELISM_VALUE.key(),argument.parallelism.toInt).
option(HoodieWriteConfig.UPSERT_PARALLELISM_VALUE.key() ,
argument.parallelism.toInt).
mode(SaveMode.Append).
save(c.getHudiConfig().basePath)
I deleted the small file, and then I can run it .
Dirty files generated by write process? I don't think it should be exposed
to reader
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
