Automated Online Backup
An online backup system becomes automated when it doesn’t need any user intervention in order to keep updated backups. With most online backup service providers, this can be achieved in just a few steps and all the user has to do is to set up the software one time, selecting paths and files to back up and the software takes care of the rest.
There are two different types of automation used in online backup systems, the first one is a time based backup, where the backup process will run at a specific time. Most system administrators or end users will set up the backup process to run overnight or at times when they know that they won’t be using any computer resources, maybe at lunch time. That way the process can start, update the backup and end without interfering with the regular use of the computer. The second type is an “idle time” automated backup, which means that the backup process will be started when there’s no computer activity, hence the term “idle time”. In order to know if a computer is idle or not, some programs will constantly monitor the CPU load or the RAM usage, others will look for any input device activity (keystrokes or mouse pointer movement), which isn’t too reliable as the user can be watching a movie but that doesn’t mean that the computer is idle. A third approach is to start with the screen saver, screen savers usually start when there’s no input activity and are disabled by some programs like DVD players, however there could be other tasks running that can slow down every process if the backup is started along with the screen saver.
A really well done backup program will look for both, CPU and RAM load, process activity and input devices to make sure that there’s absolutely no user-generated activity in the computer and therefore it’s safe to start the backup process.
Another thing to watch when automating a backup process is that it can take some bandwidth to upload the recent changes, so if you’re on a shared network you should see that backups are processed when there’s no computer activity but skipping “peak times” or those hours with more network traffic.
From the server perspective, there’s little and close to no automation, as all the server does is to remove older backup copies that won’t be used anymore. However there are some advanced backup services that will mirror your backup files to more than one server for increased redundancy, in that case, there’s more automation as each server has to copy the files and “report” to the main database to let it “know” that they’re holding a current copy.