Slow DHCP....
"Eliassen Einar (ErgoGroup)" <[email protected]>
| Newsgroups | gmane.linux.pxes.devel |
|---|---|
| Message-ID | <B719AAEE5D0F0E4EACC65EB115828147DCB3B5@SEBOX012> |
Time to wake up the list :) I'm having a weird problem on some old BayStack 310 24T switches. I'm getting almost instant messages saying "No dhcp.Network not initialized" or something like that when the client has loaded the NIC and is about to get IP via DHCP. For some reason these BayStack switches causes the DHCP request/response to take 10-20 seconds. UDHCPC is run with the -n switch which makes it fail on DHCP if an address is not received within a few seconds. (This is run in /etc/init.d/rcS) I resolved the problem by removing the -n switch. Its better to wait a little longer than to make the client halt. :) I've tried all the different port configurations on the switch. Autonegotiation ON and OFF. Half duplex and Full duplex. I've tried em all. Same slow DHCP request/response. Also tried many different NICs...tho all of them are Intel Pro 100 (various PROM versions, old ones and new) and 3Com (3C905B TX Fast also with various PROM versions, old ones and new). I've also tried updating the firmware on the switch to the current version. No change. The network is running on 10Mbit and the clients are all installed on the harddisk using Grub. The strangest thing of all....I've run thin clients on this switch since PXES 0.5-1 without this problem! I got this problem when I upgraded the clients to PXES 0.8-9 So I was wondering...has anything like NIC drivers, DHCP routines or similar been changed that might explain this ? Regards, Einar Eliassen -_____________________________________________ The contents of this e-mail are confidential to the person(s) or organisation(s) to whom it is addressed. No-one else may copy or forward all or any of it in any form without the prior written consent of the originator. If you have received this e-mail in error, please contact the sender and delete the material from any computer.