[lkml]   [1996]   [May]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectDial-on-demand in kerneld 1.3.69f.
This is something I've noticed previously but didn't get around to 
tracing it all out until this morning. I have set-up init to insure pppd
is always up and running at runlevel 5 (dedicated dial-up).

When the modem drops, pppd exits and init restarts everything pretty well
with one exception. kerneld begins to use up an extensive amount of CPU
resources and today I discovered why. When I installed 1.3.69f, I hadn't
realized it had this new Dial-on-demand feature. I found 2-3
/sbin/request-route processes running many times per second. This was
happening because I don't have an /etc/ppp/ file setup. The last
time I found kerneld not idling properly, I killed the process and my
system completely froze. I didn't try it this morning. However, when I
discovered /sbin/request-route, I edited the script and put a 'sleep
1000' in to slow it down since I don't want to bother with this
dial-on-demand stuff.

Getting to my questions:

Why does my system freeze up when I kill the kerneld process?

What options are available to completely disable dial-on-demand? Someone
on IRC suggested putting an "alias request-route off" in
/etc/conf.modules. I've done this but won't know if it works until my
modem connection dies. I could not find any mention of this alias in the
source, README's, or on the Modules homepage documents.


 \ /
  Last update: 2005-03-22 13:37    [W:0.025 / U:3.028 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site