snleee commented on a change in pull request #4464: Delete extra refreshed
segments after segment push
URL: https://github.com/apache/incubator-pinot/pull/4464#discussion_r306967874
##########
File path:
pinot-hadoop/src/main/java/org/apache/pinot/hadoop/job/SegmentTarPushJob.java
##########
@@ -48,10 +54,45 @@ public void run()
throws Exception {
FileSystem fileSystem = FileSystem.get(_conf);
try (ControllerRestApi controllerRestApi = getControllerRestApi()) {
- controllerRestApi.pushSegments(fileSystem,
getDataFilePaths(_segmentPattern));
+ // TODO: Deal with invalid prefixes in the future
+ if (_deleteExtraSegments) {
+ List<String> allSegments = controllerRestApi.getAllSegments("OFFLINE");
+ Set<String> uniqueSegmentPrefixes = new HashSet<>();
+
+ // Get all relevant segment prefixes that we are planning on pushing
+ List<Path> segmentsToPushPaths = getDataFilePaths(_segmentPattern);
+ List<String> segmentsToPushNames = segmentsToPushPaths.stream().map(s
-> s.getName()).collect(Collectors.toList());
+ for (String segmentName : segmentsToPushNames) {
+ String segmentNamePrefix = removeSequenceId(segmentName);
+ uniqueSegmentPrefixes.add(segmentNamePrefix);
+ }
+
+ List<String> relevantSegments = new ArrayList<>();
+ // Get relevant segments already pushed that we are planning on
refreshing
+ for (String segmentName : allSegments) {
+ if (uniqueSegmentPrefixes.contains(removeSequenceId(segmentName))) {
+ relevantSegments.add(segmentName);
+ }
+ }
+
+ relevantSegments.removeAll(segmentsToPushNames);
+ controllerRestApi.pushSegments(fileSystem,
getDataFilePaths(_segmentPattern));
+ controllerRestApi.deleteSegmentUris(relevantSegments);
+ } else {
+ controllerRestApi.pushSegments(fileSystem,
getDataFilePaths(_segmentPattern));
+ }
}
}
+ /**
+ * Remove trailing sequence id
+ * @param segmentName
+ * @return
+ */
+ private String removeSequenceId(String segmentName) {
+ return segmentName.replaceAll("\\d*$", "");
Review comment:
Can you add the comment on what would this regex is doing? (example will be
sufficient).
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
