Apache9 commented on a change in pull request #4202:
URL: https://github.com/apache/hbase/pull/4202#discussion_r830622759
##########
File path:
hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/storefiletracker/StoreFileListFile.java
##########
@@ -114,23 +127,96 @@ private int select(StoreFileList[] lists) {
return lists[0].getTimestamp() >= lists[1].getTimestamp() ? 0 : 1;
}
- StoreFileList load() throws IOException {
+ // file sequence id to path
+ private NavigableMap<Long, List<Path>> listFiles() throws IOException {
+ FileSystem fs = ctx.getRegionFileSystem().getFileSystem();
+ FileStatus[] statuses;
+ try {
+ statuses = fs.listStatus(trackFileDir);
+ } catch (FileNotFoundException e) {
+ LOG.debug("Track file directory {} does not exist", trackFileDir, e);
+ return Collections.emptyNavigableMap();
+ }
+ if (statuses == null || statuses.length == 0) {
+ return Collections.emptyNavigableMap();
+ }
+ TreeMap<Long, List<Path>> map = new TreeMap<>((l1, l2) ->
l2.compareTo(l1));
+ for (FileStatus status : statuses) {
+ Path file = status.getPath();
+ if (!status.isFile()) {
+ LOG.warn("Found invalid track file {}, which is not a file", file);
+ continue;
+ }
+ if (!TRACK_FILE_PATTERN.matcher(file.getName()).matches()) {
+ LOG.warn("Found invalid track file {}, skip", file);
+ continue;
+ }
+ List<String> parts =
Splitter.on(TRACK_FILE_SEPARATOR).splitToList(file.getName());
+ map.computeIfAbsent(Long.parseLong(parts.get(1)), k -> new
ArrayList<>()).add(file);
+ }
+ return map;
+ }
+
+ private void initializeTrackFiles(long seqId) {
+ trackFiles[0] = new Path(trackFileDir, TRACK_FILE_PREFIX +
TRACK_FILE_SEPARATOR + seqId);
+ trackFiles[1] = new Path(trackFileDir, TRACK_FILE_ROTATE_PREFIX +
TRACK_FILE_SEPARATOR + seqId);
+ }
+
+ private void cleanUpTrackFiles(long loadedSeqId,
+ NavigableMap<Long, List<Path>> seqId2TrackFiles) {
+ FileSystem fs = ctx.getRegionFileSystem().getFileSystem();
+ NavigableMap<Long, List<Path>> toDelete =
+ loadedSeqId >= 0 ? seqId2TrackFiles.tailMap(loadedSeqId, false) :
seqId2TrackFiles;
+ toDelete.values().stream().flatMap(l -> l.stream()).forEach(file -> {
+ ForkJoinPool.commonPool().execute(() -> {
+ try {
+ fs.delete(file, false);
+ } catch (IOException e) {
+ LOG.warn("failed to delete unused track file {}", file, e);
Review comment:
I've add a info log above which logs the loadedSeqId. Here the sequence
id for this file is contained in its name.
--
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]
