BsoBird commented on code in PR #9546:
URL: https://github.com/apache/iceberg/pull/9546#discussion_r1651099384
##########
core/src/main/java/org/apache/iceberg/hadoop/HadoopTableOperations.java:
##########
@@ -368,58 +431,63 @@ private void renameToFinal(FileSystem fs, Path src, Path
dst, int nextVersion) {
if (fs.exists(dst)) {
throw new CommitFailedException("Version %d already exists: %s",
nextVersion, dst);
}
-
- if (!fs.rename(src, dst)) {
- CommitFailedException cfe =
- new CommitFailedException("Failed to commit changes using rename:
%s", dst);
- RuntimeException re = tryDelete(src);
- if (re != null) {
- cfe.addSuppressed(re);
- }
- throw cfe;
- }
- } catch (IOException e) {
- CommitFailedException cfe =
- new CommitFailedException(e, "Failed to commit changes using rename:
%s", dst);
- RuntimeException re = tryDelete(src);
- if (re != null) {
- cfe.addSuppressed(re);
+ if (!nextVersionIsLatest(nextVersion, fs)) {
Review Comment:
@Fokko
Also, this check I admit does not solve the problem in all scenarios at the
moment, a better approach would be to call fs.rename successfully and then
check that it's commits are normal (This is the part of the code I want to
submit a PR for later.). This check just intended to fail the client as soon as
possible. Because there are a lot of clients running flink/kafka-stream to
write to flink at the same time they are running the compaction task.
Implementing fast failures is more friendly to them.
--
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: issues-unsubscr...@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org
