Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TristanM]

Hi,

I have recently setup/tested the Amazon S3 backup on my Dlink DNS-320 NAS and am a little confused as to what to expect when selecting Incremental Backup option. The text states that files on S3 will be overwritten by new files, but the incremental files/backups seem to be written to a new folder on S3 with a date suffix. So currently I have run my S3 bucket looks like this...

TestIncremental_backup << my original backup
TestIncremental_20140706102225 << subsequent backups here & below
TestIncremental_20140706102545
TestIncremental_20140706102816

These folders with timestamps only contain files that have changed and the original folder (TestIncremental_backup) has only the original files and not the new files added since first backup.

Is this correct? I was hoping/expecting a single folder which is constantly updated. Not files spread over numerous folders?

Thanks,
Tristan

[ivan]

I have to ask why you would expect anything different to what you are getting.  This is standard on professional equipment.  The idea is to allow you to go back to any version of a changed file should the need arise - think contract revisions and the need to get back to a version that was agreed.

With Incremental Backup you get an initial full backup of specified folders, after that only the files that have changed get backed up - this saves both space and time which are essential to business.

If you want everything in only one folder then you need to run a full backup with overwrite and hope you have enough bandwidth and speed to do that every three minutes (as you are doing with the Incremental Backup).

[TristanM]

Well the reason I would expect something different is that if all new files and those that have changed get copied into folder separate from the original folder it will become impossible to maintain over time. I would end up with thousands of folders each with folders existing in the initial backup and possibly new ones. If I wanted to get all my data into a single folder/location, I would have to download it and then spend a lifetime opening each folder and copy into the initial folders? Am I missing something?

TestIncremental_backup << my original backup
> folder A
   >> file A
   >> file B
> folder B
   >> file A

TestIncremental_20140706102225 << subsequent backups here & below
> folder A
   >> file C
   >> file D
   >> file E
> folder B
   >> file B


TestIncremental_20140706102545
> folder A
   >> file F
   >> file G
> folder B
   >> file C
   >> file D
> folder C
   >> file A
TestIncremental_20140706102816
> folder C
  >> file B
  >> file C
> folder D
  >> file A

How the hell would I possibly find anything if I was searching for a particular file a year later?

Thanks,
Tris

[ivan]

Sorry Tristan, it was very late at night when I posted my reply.

For us what you describe is normal practice and all document changes are tracked using software on one of our servers which means that, for example, if the contracts department wants to call up a contract that was part agreed on a specific date between specific people  they can do so by giving the date and people to the program and it will retrieve that document.

From what you are saying I have the impression that you are an individual just trying to keep your data backed up.  If that is the case then you need to use RSYNC with the relevant options set to allow overwrite of files in a specific directory.  Please be aware that if for some reason a file becomes damaged and you overwrite the good file there is no way of getting the good file back.

To be able to use RSYNC you will have to install FunPlug and set that up (have a look on the DNS-323 forum for details).

[TristanM]

Hi Ivan,

Thanks for your reply. It does look like rsync is my answer, but can I do that to Amazon S3? I have had a google around and although nothing specific says Dlink DNS-320? I see s3fs is around but not too sure if that will even help me?

Thanks again.

[ivan]

I am not sure about exactly how Amazon S3 works when trying to use external backup software but, logically, rsync should work with it without problems.

We run a private cloud setup for our clients as a remote backup service with them using rsync and everyone is happy - in one case the client does what you want while others use the system you don't want, which only shows different people want different backup strategies (the main thing is that there is a reliable backup).

When I have time I will ask the people we know if that have any experience with the Amazon cloud storage and get back here if you haven't tried it out to see if it works.