On Thu, 31 Aug 1995, Stephen Sayer wrote:
> > [STUFF DELETED]
> > This seems to be the case, except for one problem i've been having - when
> > I telnet to my Linux machine over my local eithernet, it takes about 180
> > seconds for the machine to realize i'm there. After it gives me the login
> > prompt, everything runs fine. It's just that inital delay that causes a
> > bit of a problem with some of my software that has timeouts.
> >
> > The delay happens on the SMTP port as well - it connects, but then just sits
> > there for a while before recognizing that i've established a connection with
> > a service. However, ping gets an immeadiate responce (which makes sense), so
> > I immagine it's just a problem with the telnet daemon passing the connection
> > to the service in question?
> > [STUFF DELETED]
>
> Justin,
>
> I just thought I'd drop you a line to let you know that I haven't had this sort
> of problem with any of the 1.3.x kernels including 1.3.21. I do have a local
> ether segment that I'm on however. I'm compiling with GCC 1.6.3 in a.out. I do
> wish you luck figuring this out.
>
> Later,
> <SS>
>
> {ssayer@hijinx.com | ssayer@mail.msen.com | ssayer@umcc.umich.edu}
That's intresting... I'm running a local either segment, too.. the only
diffrence from what you have stated is that i'm compiling using GCC 1.7.0..
Does anyone know if there are any known/potential problems with GCC 1.7?
I recompiled the 1.2.1 kernel last night just to see if it would work, and
it worked beautifully. Hrm.
Here's more specifics about my configuration:
IDE drivers on. (No ATAPI CD)
*NO* SCSI
Network enabled (Running 3com 3c509 16-bit ISA card/drivers)
Mitsumi CD-ROM
File Systems: EXT2, MSDOS, /proc
Profiling turned off (sorry, no time to hack the kernel! :)
So as you can see, it's pretty basic. I can't see what the problem clould
be... I've even gone as far as to disable to Mitsumi and MSDOS to see if
that solved the problem or not (to no avail).
Anyway, once again, if anyone knows anything about this problem, let me
know - it seems to be a problem with inetd spawning processes - it's just
doing it *slowly* for some reason.
- Justin -