Comments on: Why We Reboot Servers

By: Scott Alan Miller

Scott Alan Miller — Tue, 13 Mar 2018 20:15:52 +0000

James, I don’t think that your comment reflects what was in the article. We reboot so that we “know the state” of a machine, not because the application needs it. This is not related to computers resting, it is validating the correctness of the program. It’s not trusting programs simply be correct, it’s ensuring that they are and doing so at appropriate (planned) times rather than simply trusting that they are and finding out when we least expect it.

Rebooting as recommended here and any enterprise environment I’ve seen, is specifically to look for things being broken.

We can’t operate simply believing all software is perfect and will always remain perfect. We, as IT, can’t turn a blind eye to software bugs, errors, omissions and interactions. We’d be reckless to introduce more and more changes over time, without an ability to ensure that those changes work, or work together. And the more we do, the more we have to be able to back those systems out.

The entire modern world of DevOps and Cloud Computing is based on the ability to spin up and spin down, reboot at will.

By: James Frederics

James Frederics — Sun, 04 Feb 2018 17:19:16 +0000

This mindset throws away the concept of program correctness. Rebooting for updates and patches is one thing, but rebooting because “we need X program to stay running” is just subsidizing someone else’s awful program design.

Computers don’t need a rest. If you have to reboot because otherwise something quits, points to that something being broken.

By: Flensing Knife

Flensing Knife — Mon, 19 Jun 2017 01:43:11 +0000

Yep, exactly right. The same argument can also be applied to workstation rebooting. Macintosh users are particularly susceptible to the “It’s not Windows, I don’t have to reboot” syndrome.

By: lanerfy

lanerfy — Tue, 20 Dec 2016 03:10:17 +0000

oh yeah