bsync errors out on conflict

[dieterdeyke]

The following sequence crashes bsync:

#! /bin/sh

rm -rf testdir1 testdir2
mkdir testdir1 testdir2
echo xxxxx > testdir1/data
bsync testdir1 testdir2
sleep 2
rm -rf testdir1
touch testdir2/data
bsync testdir1 testdir2

[dooblem]

Hello @dieterdeyke

Can you paste the ouput you get ?

On my side this is not a crash (it just tells me that the first directory is missing):

$ sh -x ./mytest.sh 
+ rm -rf testdir1 testdir2
+ mkdir testdir1 testdir2
+ echo xxxxx
+ bsync testdir1 testdir2
Loading filelists...
Old filelist not found. Starting with empty history.

(LEFT DIR CONTENT)                   (RIGHT DIR CONTENT)             (ACTION)

data                            -->                                  (copy)

Todo in testdir1: 
Todo in testdir2: cp:1
Apply actions? [y/N] y

Applying actions...
rsync: data
Updating filelists...
Done!
+ sleep 2
+ rm -rf testdir1
+ touch testdir2/data
+ bsync testdir1 testdir2
Loading filelists...
Error: could not open directory: testdir1/ (is it created?)

[dieterdeyke]

Marc MAURICE <[email protected]> writes:

Can you paste the ouput you get ? On my side this is not a crash (it just tells me that the first directory is missing): $ sh -x ./mytest.sh + rm -rf testdir1 testdir2 + mkdir testdir1 testdir2 + echo xxxxx + bsync testdir1 testdir2 Loading filelists... Old filelist not found. Starting with empty history. (LEFT DIR CONTENT) (RIGHT DIR CONTENT) (ACTION) data --> (copy) Todo in testdir1: Todo in testdir2: cp:1 Apply actions? [y/N] y Applying actions... rsync: data Updating filelists... Done! + sleep 2 + rm -rf testdir1 + touch testdir2/data + bsync testdir1 testdir2 Loading filelists... Error: could not open directory: testdir1/ (is it created?)

I get the same output. I thought it was crashing, but it could have just been reporting the conflict in an unusual way. Never mind then. What I was really trying to do is find a test case which makes bsync really crash. That happens a lot when syncing my ext4 file system with FAT file systems on mp3 players, when I deleted files on both sides before. But so far I have not been able to make a simple example. Thanks for looking into it, -- Dieter Deyke mailto:[email protected] Get my Gnupg key: gpg --keyserver keys.gnupg.net --recv-keys B116EA20

[dooblem]

Thanks for the details.
Feel free to paste the output / exception of the error even if you are not able to reproduce it.
It may give me clues of what is going wrong.

[dieterdeyke]

Marc MAURICE <[email protected]> writes:

Feel free to paste the output / exception of the error even if you are not able to reproduce it. It may give me clues of what is going wrong.

I played some more and I think I can show the error. Attached is the script I used, and the output I got. Unmounting and remounting the DOS filesystem changes the timestamps, because FAT has only a 2 second resolution, and cannot preserve odd seconds.

…

-- Dieter Deyke mailto:[email protected] Get my Gnupg key: gpg --keyserver keys.gnupg.net --recv-keys B116EA20

[dieterdeyke]

Just saw that github does not recognize email attachments, so I will attach those files here:

test-bsync.txt
typescript.txt

[dooblem]

Thanks ! Amazing work @dieterdeyke .
After remounting the dos fs, the file time seems rounded.
I'm checking to fix this.

[dooblem]

I first fixed a bug when showing conflict of a file in vfat. As vfat files do not have permissions I should not try to display them.

Now it's not crashing anymore, but I still need to find a way to work around the vfat time problem, as it wrongly detect changes.

I will probably round the file time when on vfat, but I still need a way to detect if a directory is on vfat or not. I may use df for that but I don't think this is really portable (does it work on mac?)...

[dieterdeyke]

Marc MAURICE <[email protected]> writes:

Now it's not crashing anymore, but I still need to find a way to work around the vfat time problem, as it wrongly detect changes. I will probably round the file time when on vfat, but I still need a way to detect if a directory is on vfat or not. I may use df for that but I don't think this is really portable (does it work on mac?)...

I wrote a little function which tries to guess time resolution by looking at the already existing files. It was just a quick hack to show the concept. #! /usr/bin/env python3 # Author: Dieter Deyke <[email protected]> # Time-stamp: <2019-03-06 07:37:18 deyke> import os import sys def get_time_resolution(directory): bitmask = 0 for dirpath, dirnames, filenames in os.walk(directory): for name in dirnames + filenames: path = os.path.join(dirpath, name) statinfo = os.stat(path) mtime = int(statinfo.st_mtime) bitmask |= mtime resolution = 1 while (bitmask & 1) == 0: resolution *= 2 bitmask >>= 1 return resolution print(get_time_resolution(sys.argv[1]))

…

-- Dieter Deyke mailto:[email protected] Get my Gnupg key: gpg --keyserver keys.gnupg.net --recv-keys B116EA20