hgromer commented on code in PR #6667:
URL: https://github.com/apache/hbase/pull/6667#discussion_r1946602946
##########
hbase-backup/src/main/java/org/apache/hadoop/hbase/backup/mapreduce/MapReduceRestoreToOriginalSplitsJob.java:
##########
@@ -45,7 +45,9 @@ public void run(Path[] dirPaths, TableName[] fromTables, Path
restoreRootDir,
// We are using the files from the snapshot. We should copy them rather
than move them over
conf.setBoolean(BulkLoadHFiles.ALWAYS_COPY_FILES, true);
- FileSystem fs = FileSystem.get(conf);
+ Path backupRootDir = new Path(conf.get(RestoreJob.BACKUP_ROOT_PATH_KEY));
+
+ FileSystem fs = backupRootDir.getFileSystem(conf);
Review Comment:
the `MapReduceRestoreJob` doesn't have this issue. It delegates the files
out to a MR job which re-writes the splits, then bulkloads from the output. The
MR must be smart enough to pull from the correct file system.
as a safety measure, I've extended the test I added to restore using new
splits from a remote FS, and confirmed that it works
--
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]
