A Digital Data Backup Philosophy

<| Back to index

Why and how to do data backups
Data backups are essential to have a certainty that the data is not lost until it is required to be accessed later. A computer has different hardware and software components that are in interplay with each other and can have defects in their implementation that are internal to one component or at the interaction between two or more components. Also there exists the possibility of mistaken interaction of the user with the computer system that leads to data loss. The mitigations that together make data loss excedingly unlikely are saving often, backing locally to two different devices, backing somewhere offsite and testing the backups regularly.

Hardware and software failure
Components in the computer system are prone to failure. The first reason is the complex and unreliable nature of the technology that powers them. Hardware components that are to be sloted in a computer have built in from their design the possibility of failure in their normal operation parameters. By comparison software components are worse when it comes to reliable operation. The corectness of their functionality in unusual usage scenarios is often neglected in the design phase and relegated to fixing bugs that are reported by end users in their day to day usage of the suposedly finished product. The second reason is the market realities that govern the computer industry and that drive companies to cut corners. Competition and greed makes companies try to reduce production costs to produce a cheaper product by comparison to the competitors.

User's own mistakes
Temporary lack of attention can lead to data loss in the usage of a computer system. For example one may try to copy one file from a folder to another folder and by mistake overwrite an existing file in the destination folder. There is also the possibility of having a wrong mental model of the way the computer system actually works. For example one may think that the system will do automatical backups of data because some other system they used in the past did automatical backups of data, but the new system may require the manual operation of a user interface menu to do the backup.

Save often
One way to avoid data loss is to save the data to storage that is meant to be permanent often. For example I save roughly every time I end typing a sentence after I input the full stop mark. This avoids the situation where I lose data if there is a power loss that wipes the temporary storage area. All modern computer systems have a temporary storage area. Also most software that is manipulating data that needs to be saved to permanent storage after being edited in doesn't save the data automatically but keeps it in the temporary storage area. The existence of temporary and permanent storage areas is by design and habit because temporary storage is much faster and it doesn't degrade in time under heavy usage. Therefore it is important to learn how to save data in a computer system and the software programs you use in that system and do it often.

Use a file snapshot system
All important computer operating systems have available what is commonly called a snapshot system for files. The system litarally makes copies of the files that are being edited by users or modified automatically by software programs in a way that is efficient technologically. The system usually has to be activated to work. It will make copies at regular intervals that are spaced out in time for saving your work to permanent storage and it is useful if one may want to recover work that was drafted and saved to permanent storage but overwritten afterwards by a later save to permanent storage. Find out about the file snapshot system for files for your personal computer system and use it.

Backup to two local devices and offsite
At a more spaced out interval I save my work to two different local devices and offsite. The reason I backup to two devices is that one of those devices may fail exactly when the main device failed as well and therefore I still have one local device that holds my data. Backing up offsite to a cloud storage offering as well is important because of the possibility of disasters. When a disaster happens you cannot rely on any local device to maintain the data they were holding. Therefore local storage is not enough to ensure the permanency of your data.

Test backed up data at regular intervals
The way you can loose data on your main system is also the same way data on your backup systems can be lost. But the issue is problematical in the sense that you do not interact with the backup system every day to notice that the data has been actually lost. Therefore there needs to be a preferably automatical way to check that the data is still in place. You do not want to end up at some point in the future with a situation where your last backup system works as intended but the data it holds was written to it in a corrupted state in the first place. Check the data on the backups to see that it actually is there as intended to be.