90K views · Feb 14, 2010 · 56:19 · 4 HN recommenders · 6 mentions · discussed 2013–2023
Andrew Tanenbaum's FOSDEM 2010 talk on MINIX 3, a microkernel OS built for high reliability and security. It shows how the system detects and recovers from device driver crashes, and addresses microkernel performance concerns.
Why Hacker News points people here: HN points to it as an accessible explanation of microkernel isolation and fault-tolerant design, and as Tanenbaum's answer to performance objections.
“it's a very interesting but accessible talk for those interested in lower-level systems and fault-tolerant architectures.”HN comment, 2013
“it's a very interesting but accessible talk for those interested in lower-level systems and fault-tolerant architectures.”
Interesting, accessible talk on Minix 3 architecture
Andrew Tanenbaum presenting on Minix 3's architecture: this video . Regardless of one's opinions on microkernels vs. monolithic kernels, it's a very interesting but accessible talk for those interested in lower-level systems and fault-tolerant architectures.
“Andrew Tanenbaum explains it well”
Tanenbaum explains how driver isolation limits security exploits
It's also for security. If you find a security vulnerability in e.g. the sound driver, you're not going to get root access. Andrew Tanenbaum explains it well: this video
“Here's Andy Tannenbaum's response to the performance issues of microkernels”
Tanenbaum answers microkernel performance concerns near 47:15
Here's Andy Tannenbaum's response to the performance issues of microkernels: this video (should start at about 47:15)
“You should take a look at microkernels and the work Andrew Tanenbaum did with Minix 3”
Suggested as a modular OS design direction
Writing OS's is hard and there are no shortcuts. You should take a look at microkernels and the work Andrew Tanenbaum did with Minix 3 [1] and his modular design as this is the sort of direction you're probably looking for with customisable OS's. [1] this video
“there's nothing substantially different in this talk.”
Earlier 2010 version of the same talk
(2015) Previously: link And see [1], my note about this talk being more or less the same as the 2010 Minix talk at FOSDEM[2]. If you saw that one, or any other talk by Tanenbaum of that period, I'll spoil it for you: there's nothing substantially different in this talk. 1. link 2. this video
“Minix 3: Microkernel-based, self-healing, POSIX OS”
One entry in a list of reliable-system examples
Here's you some examples to learn from. Most are old enough that patents have expired: Why Do Computers Stop and What Can Be Done About It? (1985) link Their NonStop Architecture link QNX's systems are ultra-reliable, too link OpenVMS clusters' uptime was years to decades & mixed CPU ISA's link Systems that run forever and self-heal (Armstrong) link Microreboots link Minix 3: Microkernel-based, self-healing, POSIX OS this video
