From fd3ceb60e25adcf089584460b3659ff8e412d760 Mon Sep 17 00:00:00 2001 From: "https://www.google.com/accounts/o8/id?id=AItOawmu416zAYgYzbXVZAe30MiXoOWO4z6nGX8" Date: Mon, 11 Mar 2013 06:08:23 +0000 Subject: [PATCH 1/2] --- ...keeps_re-checksuming_duplicated_files.mdwn | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) create mode 100644 doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files.mdwn diff --git a/doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files.mdwn b/doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files.mdwn new file mode 100644 index 0000000000..845b48a995 --- /dev/null +++ b/doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files.mdwn @@ -0,0 +1,22 @@ +##What steps will reproduce the problem? + + mkdir test + git init + git annex init "test" + echo "test" > a + echo "test" > b + git annex add a b + git annex sync + git annex direct + git annex sync | grep add + git annex sync | grep add + +##What is the expected output? What do you see instead? + +The last two syncs shouldn't need to add or checksum anything. +Firstly, the output is very confusing because the files have already been added. +Secondly, the sync can take quite a while if you have lots of duplicates or a lot of files that are incidentally similar. + +##What version of git-annex are you using? On what operating system? + +git-annex version: 4.20130227 on Archlinux From 7f99294cd47ddca3ff3c822e980b40c60371436a Mon Sep 17 00:00:00 2001 From: "http://joeyh.name/" Date: Mon, 11 Mar 2013 06:31:10 +0000 Subject: [PATCH 2/2] Added a comment --- ...ent_1_cb10385a4f046bfe676720ded3409379._comment | 14 ++++++++++++++ 1 file changed, 14 insertions(+) create mode 100644 doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files/comment_1_cb10385a4f046bfe676720ded3409379._comment diff --git a/doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files/comment_1_cb10385a4f046bfe676720ded3409379._comment b/doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files/comment_1_cb10385a4f046bfe676720ded3409379._comment new file mode 100644 index 0000000000..6a5178bb6a --- /dev/null +++ b/doc/bugs/Direct_mode_keeps_re-checksuming_duplicated_files/comment_1_cb10385a4f046bfe676720ded3409379._comment @@ -0,0 +1,14 @@ +[[!comment format=mdwn + username="http://joeyh.name/" + nickname="joey" + subject="comment 1" + date="2013-03-11T06:31:09Z" + content=""" +It seems that, to fully fix this, direct mode will need to be changed to store multiple inode caches for each key. + +Since the files have the same content, but different inodes and/or mtimes, at least one of them is going to appear changed to sync each time. + +What currently happens is that the first is re-added, and so its inode cache gets stored, and then that means the second's inode cache no longer matches, and it's re-added, and its inode cache gets stored. Putting the inode cache right back in position to force re-adding the first again.. + +Storing multiple inode caches appears easy enough, but I have not yet worked out exactly how to clear out old inode caches. +"""]]