amogh-jahagirdar commented on code in PR #5373:
URL: https://github.com/apache/iceberg/pull/5373#discussion_r932802685
##########
spark/v3.1/spark/src/main/java/org/apache/iceberg/spark/actions/BaseDeleteOrphanFilesSparkAction.java:
##########
@@ -182,12 +192,26 @@ private DeleteOrphanFiles.Result doExecute() {
List<String> orphanFiles =
actualFileDF.join(validFileDF, joinCond,
"leftanti").as(Encoders.STRING()).collectAsList();
- Tasks.foreach(orphanFiles)
- .noRetry()
- .executeWith(deleteExecutorService)
- .suppressFailureWhenFinished()
- .onFailure((file, exc) -> LOG.warn("Failed to delete file: {}", file,
exc))
- .run(deleteFunc::accept);
+ if (batchDeletionSize > 1) {
+ Preconditions.checkArgument(
+ table.io() instanceof SupportsBulkOperations,
+ "FileIO %s does not support bulk deletion",
+ table.io().getClass().getName());
+ SupportsBulkOperations bulkFileIO = (SupportsBulkOperations) table.io();
+ List<List<String>> fileBatches = Lists.partition(orphanFiles,
batchDeletionSize);
Review Comment:
I'm thinking we leave the batching logic to the fileIO. Since the pattern
being followed for more advanced file io interaction is to have mixin
interfaces, this lets us delegate more to fileIO with "special" capabilities.
So I think it makes sense to remove batching from the action itself and just
let the fileIO take care of 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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
