Disaster Preparedness for Data

Are you prepared for a data disaster? I thought I was, but I discovered some gaps in my preparations. Consider the impact on your life if a phone, desktop computer or laptop...

• suffered a catastrophic malfunction
• got lost or stolen
• got destroyed in a flood, fire or other disaster
• got confiscated
• got infected with ransomware

Would you still have access to all of your photos and contacts, for example? 

Also consider the effect of losing access to a social media account or cloud service. This could happen due to one of the following scenarios (perhaps unlikely, but not impossible):

• Bankruptcy of the hosting company
• A flood, fire or cyber attack at the server facility
• A sudden change in terms of service
• Lost login credentials
• Malicious account takeover
• Temporary disruption of your internet service
• Action by a government authority

The 3-2-1 principle of data backups suggests that we should have 3 copies of our data, on at least 2 types of storage media, and with 1 copy stored off-site. By following this strategy, most of the scenarios described above would be a minor frustration rather than a major problem. 

As a starting point in evaluating my own preparedness, I made a comprehensive inventory of all the data I care about, something I had never done before. The items in my inventory included:

• Photos and videos
• Music
• Contacts
• Documents and records
• Selected social media posts
• Data from loved ones who have passed

Here are the weak points I discovered:

1. I only had one copy of my contacts (stored in the cloud)
2. I only had one copy of the photos I take with my phone (stored in the cloud)
3. My data was disorganized, spread over various devices, cloud services and storage media.

I have also recently been trying to reduce my dependence on large tech companies. This is due to concerns over their privacy practices, an interest in avoiding yet another subscription fee, plus the above mentioned (and perhaps underestimated) vulnerabilities. 

Thus, several alternatives come to mind for that off-site copy, such as copying the data to portable storage and storing it in...

• A separate building on my property (e.g. a detached shed or workshop)
• My car
• The home of a trusted friend or relative
• A safety deposit box
• A waterproof container buried in the yard
• A fire, flood and theft resistant storage container inside the home

The obvious downside of these alternative strategies is that updating the "off-site" copy is inconvenient. In my case, updating it once or twice annually should be sufficient. The second "local" copy can be updated more frequently, ideally using the automated backup process built into most operating systems. This is also an opportunity to consider whether any (or all) of the copies should be encrypted. 

Clearly there are circumstances where a cloud service would be practically mandatory (i.e. if there was a need more frequent backups of the remote copy, or a need for "anytime" remote access to the data from various devices). In that circumstance, it might be worth exploring privacy-focused alternatives to the major cloud services. 

Setting up a home server could be one component of a comprehensive backup system, even potentially providing remote access to data when away from home. However, if the server is located in the home, it does not fulfill the "one copy off-site" criteria. 

I also like to have the ability to quickly grab a physical copy of the data to take with me (e.g. if I'm traveling or if we have to quickly evacuate). If the primary copy of most data is stored on a laptop that always goes with you, then that's obviously sufficient. In my case, it will be a portable storage device. 

Lastly, this whole topic begs the question of whether I should convert legacy physical media to digital. Folks in my generation are likely to own thousands of non-digital photos, home videos on VHS tapes and a collection of compact disks or even cassette tapes.

To be clear, there is no way to eliminate the risk of data loss, only reduce it. My approach is far from perfect. In an effort to save a few bucks on automatic backups to cloud storage, I have created a system that is inefficient and prone to human error and forgetfulness. For me it's an acceptable trade-off, but I reserve the right to change my mind after I start digging.