cmoman's picture

Hello I have two turnkey machines that seem to be trying to access the same Amazon Backup service.

Machine One


Machine Two


Both these machines are virtual machines on the same server running VMWare ESXi or similar.

A long time ago I created an account through that accessed the Amazon S3 servers.  I had to cancel things when Amazon/Turnkey decided to start charging $10/month minimum last year.

However, I've now decided the tklbam storage could be quite useful and so I recently configured Machine Two to back up to the Turnkey/Amazon solution.  All worked well.

I would now like to backup Machine One.  However, I suspect that a few year ago I entered my user initialization string (API-Key) into the machine.  Complicating matters is that I may have entered a different  API-Key on the basis of prompt from an Amazon popup.

If I run tklbam-status I get

Machine One

TKLBAM: Backup ID #3, Updated Thu 2014-06-05 11:03

Machine Two

TKLBAM: Backup ID #1, Updated Thu 2014-06-05 10:52 (I'm happy with this one)

The difficultly for me is that have no idea where machine one is backed up too.  Coupled with the fact that the /etc/.git/ directory has simple exploded sending disk useage to 100%. The screenshots show machine two happily backed up.

Is there any tidy way I can start again with the backups on machine One and remove the bloated /etc/.git directory?



The backup just took a very long time. About 22hrs to do 13G.  It eventually appeared on the account that I expected it to.  I've got no idea why it was 13G though. It just a proxy server and limited the cache from the backup.



Jeremy Davis's picture

FWIW the old (v11.x) images had a bug in Webmin where it stored it's cache and logs in /etc (which is always included in backups). Also to compound this issue, etckeeper (which creates the /etc/.git directory as part of providing git version control of your /etc) was not configured ideally and didn't automatically run 'trach collection'...

Both of these issues should be fixed in the v12.x appliance but you can workaround them in your old v11.x appliance if you want. Have a look here regarding a Webmin workaround. See here to empty etckeeper's git storage, and/or here (same thread) to make etckeeper do a cleanup everytime you run apt...

Add new comment