[44493] in North American Network Operators' Group
RE: Followup British Telecom outage reason
daemon@ATHENA.MIT.EDU (Vadim Antonov)
Thu Nov 29 20:14:04 2001
Date: Thu, 29 Nov 2001 17:13:17 -0800 (PST)
From: Vadim Antonov <avg@exigengroup.com>
To: Joel Jaeggli <joelja@darkwing.uoregon.edu>
Cc: "'nanog@merit.edu '" <nanog@merit.edu>
In-Reply-To: <Pine.LNX.4.33.0111291118520.26186-100000@twin.uoregon.edu>
Message-ID: <Pine.LNX.4.10.10111291707200.5421-100000@arch.exigengroup.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Errors-To: owner-nanog-outgoing@merit.edu
Realtime stuff is not only about process rescheduling times.
The definition of real-time system is: a system which can guarantee
execution of tasks within specified time limits.
I've seen a real-life real-time system with guaranteed reaction time of
two hours (it was controlling irrigation water gates).
For routers, the "real-time" limits one needs is in 0.1 second-range;
making a system like that from a general-purpose OS is certainly doable.
--vadim
On Thu, 29 Nov 2001, Joel Jaeggli wrote:
>
> Um... The rtl kerenel runs the linux kernel as a pre-ementible low
> priority thread, has proveable worst case timing around 15uSec between
> assertion of interrupt and execution of the realtime handler, and is
> posix compliant.
>
> visit:
>
> http://www.rtlinux.org/
>
> and
>
> http://www.fsmlabs.com/
>
>
>
>
> On Thu, 29 Nov 2001, Youse, Chuck wrote:
>
> >
> > You'll forgive me for being cynical here, but I seriously doubt that any
> > Linux-derived operating systems could truly qualify as 'real-time'. To meet
> > the requirements for an RTOS, Linux would have to be so heavily mutated that
> > it would no longer be Linux.
> >
> > Cheers
> > Chuck
> >
> > -----Original Message-----
> > From: Patrick Greenwell
> > To: Christian Kuhtz
> > Cc: Alex Bligh; Paul Vixie; nanog@merit.edu
> > Sent: 29/11/01 07:49
> > Subject: RE: Followup British Telecom outage reason
> >
> >
> > On Wed, 28 Nov 2001, Christian Kuhtz wrote:
> >
> > >
> > > > I guess some time someone will realize routers are both
> > > > hardware, and software, and shock horror both, if done
> > > > well, can actually add value. [hint & example: compare the
> > > > scheduler on, say, Linux/FreeBSD, Windows 95 (sic),
> > > > and your favourite router OS (*); pay particular attention
> > > > to suitability for running realtime, or near realtime tasks,
> > > > where such tasks may occasionally crash or overrun their
> > > > expected timeslice; note how the best OS amongst the
> > > > bunch for this aint exactly great].
> > > >
> > > > (*) results may vary according to personal choice here.
> > >
> > > Don't use a non-realtime OS for something that you expect realtime or
> > > near-realtime OS functionality. There are specific systems to address
> > these
> > > kinds of needs with rather complicated scheduling mechanism to
> > accomodate
> > > such requirements in a sensible manner.
> > >
> > > Is IOS a realtime operating system? No. Are any of the other listed
> > OS
> > > realtime operating systems? No.
> >
> > Actually there are multiple Linux-based RTOSes.
> >
>
> --
> --------------------------------------------------------------------------
> Joel Jaeggli joelja@darkwing.uoregon.edu
> Academic User Services consult@gladstone.uoregon.edu
> PGP Key Fingerprint: 1DE9 8FCA 51FB 4195 B42A 9C32 A30D 121E
> --------------------------------------------------------------------------
> It is clear that the arm of criticism cannot replace the criticism of
> arms. Karl Marx -- Introduction to the critique of Hegel's Philosophy of
> the right, 1843.
>
>
