Re: Core dumps & restarting

Jon Lewis (jlewis@inorganic5.fdt.net)
Tue, 5 Nov 1996 11:28:14 -0500 (EST)


On 3 Nov 1996 raju@dontpanic.indiagate.com wrote:

> David> linux.wauug.org (aka wauug.erols.com), which is a
> David> nontrivial server doing FTP mirroring, mailing lists, Web
> David> serving, etc. was up 109 days under vanilla 2.0.0 until
> David> some of its memory physically went bad and had to be
> David> replaced.
>
> Impressive.

I had endor, a 64 Cyclades port P100, go 90 days with 2.0.10 before it
crashed...nothing at all logged. I'm assuming it was a Win95 ping, since
one of our own users tried the same thing a few days earlier on an already
upgraded 2.0.23+patches box. Unfortunately, endor crashed again only 3
days later running 2.0.23 with the second of Alan's big-ping patch. I've
got icmp to my terminal servers totally blocked now (I think) except from
certain "safe" addresses.

Both crashes were total hangs rather than reboots...even with the system
set to reboot on panic and softdog in use. In each case, there was no
response to network or keyboard, and several minutes of silence (unusual)
in the log files. Both required hitting the reset button.

------------------------------------------------------------------
Jon Lewis <jlewis@fdt.net> | Unsolicited commercial e-mail will
Network Administrator | be proof-read for $199/hr.
________Finger jlewis@inorganic5.fdt.net for PGP public key_______