Abacn commented on code in PR #38064:
URL: https://github.com/apache/beam/pull/38064#discussion_r3066988690
##########
sdks/java/io/snowflake/src/main/java/org/apache/beam/sdk/io/snowflake/SnowflakeIO.java:
##########
@@ -424,31 +433,26 @@ public Read<T> withQuotationMark(ValueProvider<String>
quotationMark) {
public PCollection<T> expand(PBegin input) {
checkArguments();
- PCollection<Void> emptyCollection = input.apply(Create.of((Void) null));
String tmpDirName = makeTmpDirName();
- PCollection<T> output =
- emptyCollection
- .apply(
- ParDo.of(
- new CopyIntoStageFn(
- getDataSourceProviderFn(),
- getQuery(),
- getTable(),
- getStorageIntegrationName(),
- getStagingBucketName(),
- tmpDirName,
- getSnowflakeServices(),
- getQuotationMark())))
- .apply(Reshuffle.viaRandomKey())
- .apply(FileIO.matchAll())
Review Comment:
Dataflow can do it because `FileIO.matchAll()` has a ReShuffle present by
default:
https://github.com/apache/beam/blob/6dd599cf3f6b40efd44fa8c084c82ca40353a051/sdks/java/core/src/main/java/org/apache/beam/sdk/io/FileIO.java#L734-L735
This introduces a fusion break and the downstream can be parallelized.
It sounds a Flink runner bug that significant impacted its performence.
Preferrably a fix should be done in Flink runner and bring backs what ReShuffle
intended to do (fusion break). Have you tried to not set useDataStreamForBatch
(only availlable for Flink 1.x)?
This PR essentially rewrites SnowflakeIO bounded read and would need a
closer eye (I could help for generic Java but less experience on Snowflake
connecor)
--
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]
