xushiyan commented on code in PR #8422:
URL: https://github.com/apache/hudi/pull/8422#discussion_r1164172550
##########
hudi-client/hudi-client-common/src/main/java/org/apache/hudi/client/HoodieTimelineArchiver.java:
##########
@@ -475,7 +485,24 @@ private Stream<HoodieInstant> getCommitInstantsToArchive()
throws IOException {
HoodieTimeline.compareTimestamps(s.getTimestamp(),
LESSER_THAN, instantToRetain.getTimestamp()))
.orElse(true)
);
- return instantToArchiveStream.limit(commitTimeline.countInstants() -
minInstantsToKeep);
+ List<HoodieInstant> instantsToArchive =
instantToArchiveStream.limit(commitTimeline.countInstants() -
minInstantsToKeep).collect(Collectors.toList());
+ // If cleaner is based on hours, lets ensure hudi does not archive
commits yet to cleaned by the cleaner.
+ if (config.getCleanerPolicy() ==
HoodieCleaningPolicy.KEEP_LATEST_BY_HOURS && !instantsToArchive.isEmpty()) {
Review Comment:
this is another special rule added to the codebase and increased logic
complexity. how can we design it better so that cleaner policy can fit nicely
with archiver. for e.g., cleaner policy can have an API say `canArchive()` and
archiver just use it. (illustration only, not as an formal suggestion). we need
to design better APIs rather than adding conditions to just make things work.
--
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]
