Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

At home, I have about 1TB of data needs backup and update frequently.

The new data is currently copied to one hard-drive and mirrored to other hard drives. However, this is not safe, if there is user error in the 1st step, all other backups are destroyed.

What's the best solution that it can: - have multiple backup. Any single hardware failure won't destroy the data. - user error tolerate. For accident deleting, overriding data, nothing should be lost.

( version control system can be good, but they don't work with hard-drives. There is no local copy of the data. The data are collected from different places )

There is also sensitive data that only stays in these off-line hard-drives. It is physically disconnected from the web, so hackers cannot touch those data. For example:

On-line machine A (PC): - public photos

On-line machine B (Mac): - personal document

Off-line hard-drive 1 - Sensitive data - backup (version-control?) of A - backup (version-control?) of B

Off-line hard-drive 2 - mirror (version-control?) of 1

share|improve this question

migrated from serverfault.com Oct 8 '11 at 15:21

This question came from our site for professional system and network administrators.

    
bup: github.com/apenwarr/bup is the best candidate I discovered so far. But it is extremely difficult to setup.... –  Tom Fishman Oct 8 '11 at 22:37
add comment

3 Answers

Install some backup solution. See Wikipedia for a (probably incomplete) list of backup software. Such software keeps history of your data changes (i.e. you can see that a file has been created on Monday, changed on Wednesday and deleted on Friday, and can recover both Monday and Wednesday version, assuming daily backups and that your backup data did not expire from repository), and because it can do full backup, and then incremental, with some time (week, month, you choose) between full backups, it does not require you to have 10 times your storage in backup media to keep 10 point-in-time images of your data.

Backup software allows you to define how often (and when) to perform full and incremental backups. Another parameter is how much space you can dedicate for backups, or for how long do you wan to keep the data. This depends on how much of your data changes. If you change 5% of your files weekly, then with backup store twice the size of your data you could keep a history of 20 weeks.

You can define external HDDs as backup media. I would recommend eSATA or USB3 drives, because full backups of 1TB data over USB2 would be rather time consuming. They are also easy to encrypt, which is handy if you would like to keep one disk at different location (your desk at work (subject to policies of the employer), friend's place, a safe box in a bank).

share|improve this answer
add comment

If you buy several 1TB hard drives (or smaller, depending on your budget) you can do a normal backup to 1 of the drives (per day, per week?) and rotate the drives in and out. If one drive fails, you lose zero day's work (it's on the master drive) so only if the master and the one backup drive fails, then you're set back a day. We use something similar to this, where we use Quantum TC-L42AN-EZ-B which can store 1.6 tb compressed per tape, and we run a daily cycle at night (7 tapes).

But if you're in a home environment, then you need to balance the cost of hard drives and the value of your data.

share|improve this answer
    
It is nice for office environment. I'm running a kind of "home-office". I update once a week. 7 tapes is not an option. But I can afford two backup hard-drives. ( plus the master distributed in many machines ) –  Tom Fishman Oct 8 '11 at 15:46
add comment

Plug an offline hard drive.

  1. If it is the first time to backup:

    :>git init
    
  2. Copy the files you need to backup:

    :>cp source_directory(anywhare) target_git_folder
    
  3. Commit git:

    :> git add ...
    :> git commit ...
    

Plug the second off-line hard drive and prepare to mirror the first one.

  1. If it is your first time using it:

    :> git --bare init
    

    (no working copy is necessary)

  2. Otherwise, do a pull:

    :> git pull src_git desk_git
    

Clean the working folder on each hard drive if necessary.

Check git's integrity often.

In this way,

  • Sensitive data on the offline hard drive can be mirrored
  • Multiple copies are satisfied
  • User error is tolerated
share|improve this answer
    
The drawback of this way is that you cannot restore a single folder easily. ( git cannot do that easily? ) –  Tom Fishman Oct 8 '11 at 22:58
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.