I have seen unpleasant incidents in which the Ethernet interfaces have
stopped responding to anything higher-level than a ping after about 20
days of use. I saw 2 of these on 2.4test1 using an EtherExpress100, and
one with 2.4test6 using the tulip driver with a PNIC card. Both were
uniprocessor kernels on uniprocessor hardware (one a PIII and one a
K6). In 2 of the 3 cases the machine was restored to full function just
by bringing the Ethernet interface down and back up again; in the 3rd case
this was not tried.
Does anyone have counterexamples of machines which have been busily
pushing network traffic much longer than 20 days under 2.4testx kernels?
Or has anyone else seen tired Ethernet interfaces? I suspect this might be
a general problem, but it is also possible the EtherExpress100 incidents
and the PNIC incident might be unrelated; so some extra data points would
be a great help.
---------------------------------------------------------
David C. Niemi (Reston, Virginia, USA) niemi@tuxers.net
Know the difference between the color of the wine and the
color of the glass. (Jalaluddin Rumi)
---------------------------------------------------------
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Sep 07 2000 - 21:00:24 EST