niklas9 / naktul Goto Github PK

It would be nice to be able to download the product as a compressed file, instead of having to clone with git on all my servers to install.

Prob not taking USE_MYSQL boolean into account, just the length of the database list

A script to restore from a backup, download it from AWS, insert MySQL data etc etc

https://cloud.google.com/products/cloud-storage/

To make sure the upload was successful, otherwise redo

Not sure if this is supported from AWS S3 etc, need to verify that first

If the BACKUP_MYSQL_DBS is empty (set to []), it would be a nice feature if it could backup all databases.

Cleanup of database dirs is now done after the upload is done, it could potentially be done before that.

In case that the upload files (which I've seen happen fairly often), lots of files will be left eating diskspace.

Make sure there's space left on device, the backup file times 2 basically, before continuing.. return error otherwise

The whole idea with nback is doing raw text backups, not binary backups (which have a higher probability of being corrupted). Because then there are way better tools (full VM snapshot etc).

Right now the mongodb module is doing binary backups with mongodump, should do it with mongoexport instead.

Right now the program crashes, should retry a few times before failing

Amazons RESTful API for S3 doesn't support larger files than 5GB in one PUT request, they need to be split up.
Right now nback crashes while trying to upload files larger than this size.

More info here:
http://stackoverflow.com/questions/10355941/how-can-i-copy-files-bigger-than-5-gb-in-amazon-s3

Set in a tuple, e.g. (1, 15, 25), and not only force to be the first as of now.

Could introduce a problem if a user sets it to 31 and all months doesn't have that date, should perhaps log an error if that happens and notify by email..

Include:

• MongoDB details (completely left out now)
• Include what file dirs have been added (and potentially their size)
• Better file size handling, even if the backup is 2.7GB now only '2GB' is displayed right now

This is non-existent right now.
If upload to AWS S3 fails, taring fails etc, we should send error notifications with full traceback.

Cleanup should still execute as well, so we don't end up eating more and more memory every time the backup is executed (until it's manually stopped).