Confused: s3sync not working as expected...

[hnzz]

Hi, setting up an automated backup for my linux box using s3sync. It works quite well  but i see some, in my opinion, strange things.. When i have done a backup and then rerun the backup script with just a couple of files changed, it still re-transfers all the data in my directories again with one exeption: the directory containing my html files. From the latter it only transmits the changed files in a rsync kind of style. As I like to have this to happen all the files in the filesystem i was wondering how to achieve this..

any answers very welcome..

hnzz

[lowflyinghawk]

for starters, giving the exact command line your are running would probably help in diagnosis.

[hnzz]

there you are:

this is the content of the script I use...

#!/bin/bash
cd /usr/local/s3sync/
export AWS_ACCESS_KEY_ID=xxxxxxxxxxxxxxx
export AWS_SECRET_ACCESS_KEY=xxxxxxxxxxxxxxxxxxxx
export SSL_CERT_DIR=/usr/local/s3sync/certs
ruby s3sync.rb -r --ssl -v --delete /home/backup/remote/ spip:/

[ferrix]

Can you reproduce this issue with a small set of dirs/files, and then archive them up and attach that here so I can test it on my system?  It sounds like this might be some kind of obscure comparator bug.  If so, that will be hard to diagnose remotely.

[hnzz]

Ok, i have created a directory with a smaller set of files. With this set of files s3sync performed flaweless.
I then just have deleted all the already uploaded files on S3, deleted the bucket, recreated the bucket, sync my original set of files and now it works as a charm.. I guess due to ? the file info of the orginal bucket was corrupted or what. Anyway, it works now perfectly.. and thanks for your help.

hnzz

[lowflyinghawk]

ok then, now you can claim it is self-healing software, another required buzzword checkoff ;-).