tftp: timeout

[Roland Rohde <roro@artificial-organs.de>]

Hello,

since a few months I'm running netboot 0.8.1. on four tulip chipset
adapters.
Everything seems to work fine with one exception. A menu should be
loaded after dhcp and a OS boot file after that. But in most cases the
menu needs three to six "tftp: timeout" messages until it is
transferred. Sometimes the clients even have to be restarted. If the
bootmenu is loaded and the user chooses an OS, the OS boot file is
catched nearly always at once without a timeout message.
Inspection of the messages file on the Server (suse 6.0, 2.0.36) shows,
that every tftp request is recognized. But the clients seem to timeout
before the server is able to answer. 
The main problem for me are not the impatient users. But every few weeks
the server starts to refuse tftp requests because there are about 100
forgotten tftpd processes (Gero's modified tftpd) running. Killing them
lets the sever answer again. 
Does anybody know why a server reacts fast on requests for dhcp, http,
nfs, pop, smtp, etc. but so lousy on tftp?
Is there really no mechanism that kills outdated tftpd processes so that
I have to do that manually or by a cron job?

roland

roland rohde <roro@artificial-organs.de>
===========================================================================
This Mail was sent to netboot mailing list by:
Roland Rohde <roro@artificial-organs.de>
To get help about this list, send a mail with 'help' as the only string in
it's body to majordomo@baghira.han.de. If you have problems with this list,
send a mail to netboot-owner@baghira.han.de.