Backups are T+1
Few people invest time and money to set up and maintain a reliable backup strategy for their personal data. But even then, most fail to realise what backups are good for and what not.
The important thing to have in mind is that your backup is designed to be used tomorrow, or generally speaking, a few days after today. I call this T+1: today + 1 day. If your laptop get stolen, if your hard disk fails, if you get hacked, your backup is there to help you recover, as soon as you realise the problem.
We make a number of assumptions about our backups, that are good enough when we are talking “T+1". Here are some:
- if you used an external storage device to store your backups, you will be able to plug it in to your new computer too.
- if you verified your backup today, chances are that it will be in good shape tomorrow too.
- if you password protected your backups, and you know your password today, you will know it tomorrow too.
- if you paid for a cloud backup service yesterday, your backup will be there tomorrow when you need it. You also have (or can download) the required software to use the backup, and you know your username and password.
- Your backup is intended to be used by you —obviously.
The T+10000 problem.
What if I want to store a couple of files today in a way that they will be accessible by me, or someone I trust, for decades, even if I disappeared, or went broke, or forgot about them?
I call this the T+10000 problem, where T+10000 stands for today+10000 days (that’s close to 30 years).
Unfortunately, backup solutions (T+1) are a bad fit for T+10000 for a number of reasons.
For example, storage media used today will not be usable in a couple of decades, I can not safely assume I will be able (or willing) to support financially my cloud backup solution for so many years, and if someone else (say, my kid) would like to access any cloud service they will have to know my credentials which will have changed by then… There are so many moving parts (technology, financing, expertise, time investment) that can fail one way or another in 20 or 30 years.
T+10000 is a great business opportunity.
When it comes to public data, the T+10000 problem could be addressed by services like archive.org: for example, my blog posts will be on archive.org even if I shut down my blog, or stop paying for its hosting. (And that’s one of the many good reasons to support archive.org).
But when it comes to private, or semi-private or confidential data things are much more complicated: it’s not “just” about preservation but also about security, privacy and transfer of access control rights. There are also issues related to data formats, extracting data from social networking services and deciding what to archive and what not.
There’s almost nothing available in this direction for the consumer market. Our company, Longaccess, offers a solution for a good number of use cases, but I expect to see much more going on on this field: There’s a big opportunity to create a whole industry of services and products.