Gone are the days when I used to be able to keep my Linux server up and running for upwards of 150+ days. These days updates are so frequent that I just about manage a month before the next set are released.

It’s not just my server either, the same is true of the servers I used to look after in previous jobs.  These weren’t just exclusively Windows or Linux servers either, it was a mix of both.

Some people take the view that don’t need to apply an update if you don’t use that particular bit of the system, but I’m in the opposite camp. What happens if one day you install a piece of software that uses that bit of the system that you’ve never updated…

On Windows it’s rare to apply and an update doesn’t require you to reboot after applying it, while on Linux it’s often easier as you can just restart the service that you’ve just updated.  This can lead to bigger uptime for Linux, but these days Linux systems are so complex and the kernel updates are that frequent that I tend to reboot after most major updates (unless I’m 99% sure that I don’t need to).

Getting that 100% uptime is becoming harder and harder.