From 82ce149601f196e2e727908883bfb1b5a9d37319 Mon Sep 17 00:00:00 2001 From: tedyu Date: Mon, 5 Nov 2018 18:59:44 -0800 Subject: [PATCH] HBASE-21381 Document the hadoop versions using which backup and restore feature works (liubang) --- src/main/asciidoc/_chapters/backup_restore.adoc | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/src/main/asciidoc/_chapters/backup_restore.adoc b/src/main/asciidoc/_chapters/backup_restore.adoc index fbf93777f24..592faebfc16 100644 --- a/src/main/asciidoc/_chapters/backup_restore.adoc +++ b/src/main/asciidoc/_chapters/backup_restore.adoc @@ -50,6 +50,23 @@ restore feature is ideal for creating "cold" copies of data (where a manual step Previously, users only had the ability to create full backups via the ExportSnapshot functionality. The incremental backup implementation is the novel improvement over the previous "art" provided by ExportSnapshot. +The backup and restore feature uses DistCp to transfer files between clusters . +link:https://issues.apache.org/jira/browse/HADOOP-15850[HADOOP-15850] fixes a bug where CopyCommitter#concatFileChunks +unconditionally tried to concatenate the files being DistCp'ed to target cluster (though the files are +independent) . Without the fix from +link:https://issues.apache.org/jira/browse/HADOOP-15850[HADOOP-15850] , the transfer would fail. +So the backup and restore feature need hadoop version as below + +* 2.7.x +* 2.8.x +* 2.9.2+ +* 2.10.0+ +* 3.0.4+ +* 3.1.2+ +* 3.2.0+ +* 3.3.0+ + + [[br.terminology]] == Terminology