steveloughran commented on PR #5391: URL: https://github.com/apache/hadoop/pull/5391#issuecomment-1443728823
This is interesting. I know there are deployments without hdfs around (e.g. azure clusters), but do see that the import is there for snapshot updates (hdfs only) with explicit imports of SnapshotDiffReport. tracing it back, if you use "-diff" on the command line then hdfs *must* be on the classpath. your link flags up that it is already in copymapper; looking at that I don't see it in branch-3.3; it came in with HADOOP-14254. Given it is in there on a codepath hit with the option to copy erasure policy (and skipped if not), then again, provided the change goes in such that it is optional, your patch isn't going to force in a new run-time dependency, is it? let me look at the new patch some more without worrying about that detail...the builder API is public. Be aware that we are always very nervous about touching distcp because it is fairly old and brittle code that is used incredibly broadly -not just on the command line but actually at the Java API from applications like hive. I think this is fairly low risk but will highlight the JIRA on the HDFS mailing list so they can review it to. -- 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]
