Thursday, October 1, 2009

Backing up your data

This is always a touchy subject. Most people, even long time technologists, do not bother backing up their data until far too late. Some years back, we learned the hard way how vital regular backups are when one of the disks failed on a home built PC. $2200 to a cleanroom later, we had all of that data back.

There are plenty of fantastic data backup programs and most large external drives come with backup software installed on the device. Mac OS and Windows deliver backup software with their operating systems bundled for free. Alternatively, you could buy an external backup server specific for your environment (computers) and run it on your internal network. Finally, you could choose an online backup company (fashionably: backing up to the cloud) and enjoy automatic, remote backups.

So, what is a consumer to do? The long and short of it is, I don't think there's one golden solution out there for everyone. Solutions depend upon your needs, how important your data is to you (and therefore how much you want to spend on backups), how technical you are or what you are willing to pay for setup, how paranoid you are and what level of risk you wish to tolerate.

There are no end of configurations and knobs you can twist when setting up these things. Doing backups? Should they be incremental, differential, full copy, replaced? What do those mean and why should you care?

What about a network storage (NAS) device for backups? Will it support Mac and Windows backups (Time Machine and Windows NTFS)? Do I have to purchase extra software to remote mount my NAS so that my other OS can "see" it? Also, how should you format your NAS's disks? RAID 0, 1, 5, 10, 01? Again, what do those mean and why do you care? Can the NAS disks be partitioned so that some backup data is only RAID 0?

Time Machine only works for the same computer; you need to dig deep and extract the files in a non-standard way if your computer is stolen and replaced (v. adding a replacement disk drive to the same computer). Are you backups working?

Will your chosen cloud computing backup provider go out of business and what happens to your data? Is it secure? Can you trust your financial data to be stored there? Is it encrypted? Who holds the encryption key(s)?

What is your offsite data backup plan? Do data copies go to local places or remote? What if an earthquake hits? Can your business recover? Will it matter?

It's enough to drive a sane person mad.

What's my backup plan?

For now, I use Time Machine from both of my computers. I keep a small number of files in Google Docs. All email is either POP (Yahoo) or IMAP (Google). All data in Parallels resides in Mac file hierarchy so that it can be backed up, except for Quicken data because Quicken does not play well in that configuration. However, Quicken backups go to Mac file hierarchy.

This is not enough, but it's a start.

What I would like:
A NAS that can handle Time Machine and Windows backups, RAID 10 (not 5), autobackup from NAS to the cloud, iTunes data not backed up or not backed up duplicate (as it resides on two computers and the NAS), photos backed up and available from NAS as a webserver to the world.

The cloud backup folks would want me to just use their services, but what if they go out of business? The NAS folks don't see the need for a cloud backup, but I don't want to be bothered with roll your own offsite backups/storage.

There's one NAS that almost fits the bill, but it's RAID 5 (no RAID 10): Synology Disk Station DS409+. It has almost all of these things, plus SVN (source control) and allows third-party application development. A most excellent offering. I'm concerned about the lack of RAID 10 as I'm one of those folks who thinks RAID 5 is not worth it. However, if you are going to do RAID 5, be sure to pick up disk drives from different batches, otherwise, they might all fail similarly and due similar problems during manufacturing.