Can't ping, though DHCP returns IP

[heckflosse230]

Hi,

I'm a Windows web developer thrown off the deep end to develop a utility on an embedded Linux board.

I'm having issues with network connectivity.

Last week I was able to get network connectivity on the Freescale MPC8313 board running Linux version 2.6.20.

Eventually, I would like to get the MPC8313 board running as a NFS client connected to a SUSE installation in VirtualBox running on my Windows development machine.

I got the PC, virtual SUSE installation and the board to ping each
other happily.

Now however, the Linux board is having connectivity problems and I haven't made any changes. I have the same issue at home and at work.

I can't ping in or out of the board.

How do I diagnose this?

Thanks,
Chris Plasun

[PastAsNew]

I'm no expert here, but would you be better to use fixed IP addresses till you track it down, to remove one possible variable from the loop? Or alternatively have you got some fixed and some DHCP and ended up with an address clash? But I'm not strong on network stuff.

Like you I'm new to Linux, in my case migrating from embedded VxWorks.

[heckflosse230]

I'm no expert here, but would you be better to use fixed IP addresses till you track it down, to remove one possible variable from the loop? Or alternatively have you got some fixed and some DHCP and ended up with an address clash? But I'm not strong on network stuff.

Like you I'm new to Linux, in my case migrating from embedded VxWorks.

Hi,

Yes, I've tried static but still I get issues. I'm going to have install tshark or tcpdump and see what's going.

It's nice that there are other Linux noobs to commiserate with! =)

Chris

[heckflosse230]

Just in case anyone finds this useful, I found the solution to this "problem".

eth0 is not active and eth1 is the active interface and while I don't really understand how the IP routing table is used, the internal connectivity problem occurs if :
- eth0 is up
- eth0 subnet mask is the same as eth1
- eth0 is defined first, above eth1

If eth0 has a different subnet mask then there is no problem.

I systematically reproduce the issue and so it looks like problem is solved!

I hope this helps someone in the future.

Chris Plasun