Test Driving GNU Hurd, With Benchmarks Against Linux

[JanC]

I didn't mean to say tests inside a VM are completely useless, but they are meaningless without knowing what is being tested. Doing the same tests on the same machine on bare hardware might help, provided the machine is booted with only one core/thread active. And of course, as somebody already mentioned, more different tests would be useful too.

[vermaden]

Benchmarking OSes under virtual machines reminds me this

http://www.robertgering.de/wp-content/uploads/WindowsInception.jpg

[Adarion]

Realistically, BSD was just emerging from its legal difficulties at the time, and would have been the "canonical free OS" if Linux had not come along. So we already have TWO free OS with broad hardware support thanks to Linus and Bill Joy.

Well yes, but BSD has in my eyes this one way license problem. Linux is GPL/LGPL and this is free and stays free. I guess that also attracts more developers to it than to a BSD licensed project. Some see it as freedom to have the ability to even close / relicense the code under a different license but I don't.


PS on the article from Michael: It's indeed sad that the tests were carried out in a VM, that seems not too really realistic to me. At least I expect all sorts of falsifying effects from the VM/host machine. But with this hardware support projects like HURD, Plan0/Inferno might stay an experimental research project for geeks. But maybe some of the interesting info won there could be used in the more mainstream products.

[yogi_berra]

and how much are these benchmarks worth - running in a vm?

how much are these benchmarks worth - when one of them can't even do SMP?

What, not dropping everything to switch to the hurd based on a phoronix article?

[energyman]

and how much are these benchmarks worth - running in a vm?

how much are these benchmarks worth - when one of them can't even do SMP?

[Delgarde]

I don't know about the deep things but I'm surprised that Hurd is actually comparable to linux.

Not that surprising. The handful of tests Hurd could actually run seem to be the kind that just max out the CPU - so as long as the kernel doesn't impose too much overhead, I'd expect similar performance from any OS.

[JanC]

I wonder how reliable tests run inside a VM are, e.g. some of those tests might well be testing the speed of I/O emulation in Qemu instead of the speed of the kernel running on it...?

[frantaylor]

I'm glad Linus chose monolithic approach so we already have a free OS with a broad hardware support today.

Realistically, BSD was just emerging from its legal difficulties at the time, and would have been the "canonical free OS" if Linux had not come along. So we already have TWO free OS with broad hardware support thanks to Linus and Bill Joy.

[Adarion]

Thanks for the insight into a recent Hurd software compilation. Was an interesting read for me.

Still it lacks devs and microkernels are a fine thing in theory but need lots of work in reality.
I'm glad Linus chose monolithic approach so we already have a free OS with a broad hardware support today.

[frantaylor]

I think it's interesting that, despite the vocal Microkernelist group extolling the superiority of their championed design methodology, these few tests show a system only managing to be just as performant.

Considering the relative maturity, this isn't unexpected. But if you consider the years (~7) that have gone into false starts (>=2) at moving to a different design for said microkernel.... well, I can't help but feel as though the point was missed somewhere.

Umm, you do know that these benchmarks are POSIX programs and as such are running in "POSIX compatibility" mode in the hurd? You aren't going to see any microkernel performance enhancements until we have programs that are written for that environment.

And actually it's really pretty impressive that their "compatibility mode" is running almost as well as Linux.