strange DHCP server problem on SuSE 10.0
I have a SuSE 10.0 machine running the DHCP server (I upgraded the machine from SuSE 9.3 a couple of weeks ago). The DHCP server still seems to work perfectly well with all my machines except for one (a 233MHz PII machine with a realtek 8139 NIC). For all the working machines, I see the normal 'discover offer request ack' sequence in the log files - e.g.: Nov 1 20:42:47 opteron dhcpd: DHCPDISCOVER from 00:c0:26:75:ee:89 via eth0 Nov 1 20:42:47 opteron dhcpd: DHCPOFFER on 192.168.2.220 to 00:c0:26:75:ee:89 via eth0 Nov 1 20:42:47 opteron dhcpd: Wrote 2 leases to leases file. Nov 1 20:42:47 opteron dhcpd: DHCPREQUEST for 192.168.2.220 (192.168.2.20) from 00:c0:26:75:ee:89 via eth0 Nov 1 20:42:47 opteron dhcpd: DHCPACK on 192.168.2.220 to 00:c0:26:75:ee:89 via eth0 but on the non-working machine, there's a discover and an offer, but no DHCPREQUEST comes back from the client: Nov 1 20:34:57 opteron dhcpd: DHCPDISCOVER from 00:06:4f:05:84:a4 via eth0 Nov 1 20:34:58 opteron dhcpd: DHCPOFFER on 192.168.2.220 to 00:06:4f:05:84:a4 via eth0 The /var/log/messages show that the 'now borken' machine was working perfectly up to the upgrade to SuSE 10, and hasn't worked since. I suppose there's a slight chance that the non-working machine broke in some way at exactly the same time as I upgraded to 10.0, but there's not a lot to go wrong as it's a thin client machine using etherboot. My dhcp.conf file remains exactly as it was before the upgrade. Anyone got any ideas what might be wrong? Thanks -- Phil Driscoll
On Wednesday 02 November 2005 08:23, Phil Driscoll wrote:
I have a SuSE 10.0 machine running the DHCP server (I upgraded the machine from SuSE 9.3 a couple of weeks ago).
The DHCP server still seems to work perfectly well with all my machines except for one (a 233MHz PII machine with a realtek 8139 NIC).
For all the working machines, I see the normal 'discover offer request ack' sequence in the log files - e.g.:
but on the non-working machine, there's a discover and an offer, but no DHCPREQUEST comes back from the client:
Nov 1 20:34:57 opteron dhcpd: DHCPDISCOVER from 00:06:4f:05:84:a4 via eth0 Nov 1 20:34:58 opteron dhcpd: DHCPOFFER on 192.168.2.220 to 00:06:4f:05:84:a4 via eth0
I used the same dhcpd.conf file on a 9.3 machine and my 'broken' machine started working again. Next, I ran ethereal on my 10.0 and 9.3 machines to see if I could spot any difference between the DHCPOFFER messages. On the 10.0 machine the DHCPOFFER message had a Next server IP address of 0.0.0.0, whereas the 9.3 machine set this value to the IP address of the 9.3 machine. So, on the 10.0 machine I added a next-server directive to the dhcp.conf file setting the value to the IP address of the 10.0 machine. Now, by previously 'broken' client responds to the DHCPOFFER and boots up as an LTSP thin client. I hope this helps anyone who finds themselves with the same problem. -- Phil Driscoll
participants (1)
-
Phil Driscoll