[erlang-questions] Millions of processes?

Bengt Kleberg <>
Fri Sep 26 07:36:25 CEST 2008


Greetings,

I had not at the time heard of a generic benchmark for this kind of
comparison. There was no benchmark of anything mentioned in the article.

The benchmarks I found, with micro kernels being faster than macro
kernels, involved running a web server, copying file trees, etc.

I do agree that some things are more important than performance (slow
and correct is generally better than fast and wrong), but this is (as
usual) not a black and white question.
What if we get an almost correct answer in time, compared to an accurate
answer too late?


bengt

On Thu, 2008-09-25 at 10:30 -0400, Toby Thain wrote:
> On 25-Sep-08, at 2:39 AM, Bengt Kleberg wrote:
> 
> > There where other micro kernels around at the time. At least 2 had
> > benchmarks showing them to be faster than ''macro kernels''.
> 
> Does a generic 'faster' metric/scale really exist when comparing  
> operating systems?
> 
> What I tend to see in benchmarks of complex systems like filesystems,  
> databases, operating systems is that it's multivariate, workload  
> dependent, and simply leads to differing sweet spots. And frequently,  
> there are important attributes unconnected with performance...
> 
> --Toby
> 
> >
> >
> > bengt
> >
> > On Thu, 2008-09-25 at 15:20 +1000, jm wrote:
> >> This was generally true only of the early micro kernels.  
> >> Apparently, the
> >> Mach kernel is slow when compared to the newer L4 kernel. see
> >> http://en.wikipedia.org/wiki/L4_microkernel_family#History
> >>
> >> Jeff.
> >>
> >> Bengt Kleberg wrote:
> >>> Greetings,
> >>>
> >>> It is not only _ideas_ in computer science that are held in  
> >>> contempt (as
> >>> quoted below). Ages ago (before 2000) I read an article about  
> >>> Linux in
> >>> embedded environments. The article quoted Linus Torvalds on why  
> >>> not to
> >>> use micro kernels. The reasons where that they are:
> >>> 1 Experimental
> >>> 2 Complex
> >>> 3 Slow
> >>>
> >>> After looking around for a while I found plenty of articles about
> >>> commercial micro kernels, and benchmarks showing micro kernels  
> >>> running
> >>> workloads faster than monolithic kernels. So 1 and 3 seemed to be
> >>> incorrect.
> >>> I submitted these findings to the magazine, which prompted an answer
> >>> from Mr Torvalds. He assert that all three where true, but did not
> >>> discuss what I had found. So (IMHO) it is also facts that are not  
> >>> held
> >>> in very high regard.
> >>>
> >>>
> >> _______________________________________________
> >> erlang-questions mailing list
> >> 
> >> http://www.erlang.org/mailman/listinfo/erlang-questions
> >
> > _______________________________________________
> > erlang-questions mailing list
> > 
> > http://www.erlang.org/mailman/listinfo/erlang-questions
> 




More information about the erlang-questions mailing list