I´m working on an experimental OS and I´m unsure whether choosing a modular monolithic kernel over a microkernel was the right architectual trade-off. I´d like technical feedback on that decision. #213646
Unanswered
zeropatience-dev
asked this question in
Programming Help
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
🏷️ Discussion Type
Question
Body
I'm working on an experimental operating system and made a deliberate architecture decision that I'd like to get some feedback on.
I chose a modular monolithic kernel over a microkernel.
The main reason was simplicity.
A microkernel was attractive because of the stronger isolation between components, but it also introduces additional complexity: IPC boundaries, service lifecycles, synchronization, timeout handling, restart paths and more distributed failure modes.
For a hobby OS, I decided that reducing this complexity was more valuable at the current stage.
The idea is not to build one large unstructured kernel. The kernel is intended to remain modular, with clear responsibilities, internal APIs and testable boundaries between components.
My assumption is that stronger isolation boundaries can still be introduced selectively later, where the security benefit justifies the additional complexity.
So the trade-off I made was essentially:
less isolation now in exchange for lower architectural and implementation complexity.
What I'm unsure about is whether this is a reasonable trade-off, or whether I'm creating technical debt that will be difficult to undo later.
I'm especially interested in feedback on:
I'm not looking for a simple "microkernel vs. monolith" verdict. I'm mainly interested in the trade-offs and in assumptions I may have missed.
You definitely don't need to answer all of these — even feedback on a single point would be really appreciated.
Guidelines
All reactions