Alexander Hall wrote:
From reading the fabulous source (/usr/src/usr.sbin/inetd/inetd.c; look
for "toomany"), I can only conclude that they (-R and [.max]) indeed
work as I had guessed, i.e. -R changes the default from 256 to whatever
and that the [.max] suffix allows you to specify it for each specific
service, if needed.
Are you sure that there is no other service that is affected by the -R
switch that you had not thought of?
Fore that, I am sure as a default for me on all my servers is to simply
not run inetd what so ever. I don't use it for anything at all, except
here in this case for the tftpd only.
man(8) inetd does really indicate this as well as I saw before:
"The op-tional ``max'' suffix (separated from ``wait'' or ``nowait'' by
a dot) specifies the maximum number of server instances that may be
spawned from inetd within an interval of 60 seconds. When omitted,
``max'' defaults to 256."
I will need to dig more then.
My problem is really that I do have the wait.6000 in the inetd.conf file
as provided earlier, but I get lots of deny connection when a group of
VoIP Cisco phones are reset at once, witch happen often, but I do not
get this when I left that in the inetd.conf with the wait.6000, but I
also manually start inetd -R 1024, witch in itself would be
contradictory to the wait.6000 here obviously.
That's really the problem I am dealing with.
So, only wait.6000 in inetd.conf doesn't fix the problem if I do not
also start inetd -R 1024. Weird then based on the man page.
Best,
Daniel
