Tom Peters wrote:
No route. Just a crossover Cat5.
The Linux box has an IP address. The other box has one. If they're on the same subnet, there is still a route between them, it's the one that says "go straight out the wire." If they're not on the same subnet, it has to know what the correct default gateway address is. If it's not configured the same as the Windows box was, it probably won't work.
Tried SuSE with the same IP as the XP box, and a few others in the same range. Didn't make any apparent difference.
What static IP? Same one as the Windows box you tried? Same IP with a different MAC address means screwed up ARP table, possible confusion. Not supposed to matter on a network with no layer 3 involvement. Not SUPPOSED to matter... Try another address on the same subnet.
Previous experience (mass hardware swapping) suggests that the industrial unit involved is pretty immune to that type of thing. But yes, I have tried multiple IPs.
On Fri, 2005-12-02 at 11:16, David McMillan wrote:
Tom Peters wrote:
No route. Just a crossover Cat5.
The Linux box has an IP address. The other box has one. If they're on the same subnet, there is still a route between them, it's the one that says "go straight out the wire." If they're not on the same subnet, it has to know what the correct default gateway address is. If it's not configured the same as the Windows box was, it probably won't work.
Tried SuSE with the same IP as the XP box, and a few others in the same range. Didn't make any apparent difference.
What static IP? Same one as the Windows box you tried? Same IP with a different MAC address means screwed up ARP table, possible confusion. Not supposed to matter on a network with no layer 3 involvement. Not SUPPOSED to matter... Try another address on the same subnet.
Previous experience (mass hardware swapping) suggests that the industrial unit involved is pretty immune to that type of thing. But yes, I have tried multiple IPs.
Earlier someone suggested the problem might be the difference in ping
packet size between windows and Linux. An experiment to test that would
be 'ping -s 32
Earlier someone suggested the problem might be the difference in ping packet size between windows and Linux. An experiment to test that would be 'ping -s 32
' to product a ping the same size as windows.
Must've missed that one, or my email is arriving asynchronously. I'll try that when I'm back in the shop on Monday.
Jim Cunning
participants (3)
-
David McMillan -
Jim Cunning -
skyefire＠skyefire.org