DSL Embedded :: DHCP client (pump) catching strange IP address



Here are the details:

WinXP laptop (Dell D600) running Windows XP.  Uses Broadcom ethernet adaptor and Dell TrueMobile Wireless adapter.

Everything was working fine for a couple of days, I was able to get on the network and browse with Dillo and Firefox and even use Rdesktop at a client's shop to admin some Windows servers.
When DSL boots up, everything is normal.  When I FIRST ran DSL embedded it was unable to find the correct driver for the broadcom adapter so I modprobed ne2k-pci and was able to renew the adapter and catch an IP address, allowing me to use the network no problems.

Today when I boot up DSL embedded, I no longer have t modprobe for the driver...it find it on boot, and pump seems to grab an IP address with no problem.  The problem is it's catching an IP address of 10.0.2.15/24 and adding a default gateway of 10.0.2.2 and a DNS server in /etc/resolv.conf of 10.0.2.3  
I can not surf the web of do anything network related because that IP address scheme is not from my DHCP server.  My small LAN is setup using an IP address scheme of 192.168.1.100/24 with a default gateway of 192.168.1.101.   This is a pretty standard configuration and my other 5 machines are happy with this, including the WinXP host, which is catching 192.168.1.113 as it should an pointing to the appropriate gateway, 192.168.1.101

I've tried releasing and renewing several times and it catches the same IP address, 10.0.2.15 with the same gateway, 10.0.2.2
I've tried modprobe -r ne2k-pci and reloading the module and then configuring the ethernet interface manually as so:
# ifconfig eth0 192.168.1.121  
The adapter will take the IP address I specify so I was happy, I then configured the default gateway as follows:
#ifconfig add default gw 192.168.1.101
wonderful!  when I list the routing table it even finds the default gateway.
However I can not ping anything...including machines on my own network...and I can't ping the default gateway.  Something is radically wrong.

I went a step further and took the .zip file I downloaded and extracted DSL-2.0 embedded into another folder and booted it up "fresh".  It booted without incident and went on to find an IP address and default gateway, 10.0.2.14 with the gateway at 10.0.2.2

Anyone else having this problem?  I've only had this running for a few days and DAY ONE was prefect....I was able to catch an IP at 3 different client locations an catch IP addresses on their networks with no problems.  I was showing it off to my client's admins who were in awe.  Now all of the sudden the networking seems to have shit the bed. Not sure why.  I've used pump before in past lives and it's fairly reliable IMO, but I don't really think that's the problem.

One final last word - a friend of mine in my LUG also downloaded DSL the same day as I did...and he too also had good luck with it the first day. He is now experiencing the same exact problem, he's getting the same exact IP address (10.0.2.15) and the same exact gateway (10.0.2.2) and DNS server (10.0.2.3)

Anyone who has experienced this, I'd appreciate hearing from....I'd like to work this out...I'm sure it must be a minor bug or config issue.  I'll keep playing with it and post back if I come up with any significant finds.

smed
It is completely normal because you are running a virtual machine.
See also http://damnsmalllinux.org/cgi-bin....hl=ping
Quote (roberts @ Dec. 09 2005,19:27)
It is completely normal because you are running a virtual machine.
See also http://damnsmalllinux.org/cgi-bin....hl=ping

It's may be normal, but it's non-functional.
I don't care how the VM does it's thing, but I can't ping anything on the Internet or inside my LAN.  Traceroutes fail completely after the hop to the host machine.  The nameserver that the DHCP client writes to /etc/resolv.conf doesn't resolve *anything*, and the beside being able to ping the default gateway, it doesn't seem to be routing any traffic, which is pretty much the gateway's job.

There is nothing "normal" about that....the network is no more functional than if the interface didn't have an IP address and the routing table was blank.
I am not going to argue with you... Your post is nothing new.
10..0.2.15 is the standard response from Qemu virtual machine.
If your windows networking is functional with YOUR specifics then Qemu's default network will pass thru using 10.0.2.x
Please take your questions regarding Qemu's networking to the Qemu support site.
Quote (roberts @ Dec. 09 2005,20:49)
I am not going to argue with you... Your post is nothing new.
10..0.2.15 is the standard response from Qemu virtual machine.
If your windows networking is functional with YOUR specifics then Qemu's default network will pass thru using 10.0.2.x
Please take your questions regarding Qemu's networking to the Qemu support site.

ok here's an update.
Normal network debugging tools such as ping and traceroute do not work as expect through QEMU.

I CAN indeed load web pages with the seemingly erroneous IP address and gateway assignment, but I can't resolv any names for some reason...and the odd thing is that it WAS working the first day I tried this.

I've noticed now that running DSL embedded on WinXP, Win2K and Linux all seem to behave the same in regards to the way it handles network functionality.  I'm surprised there aren't more posts regarding this issue.  Maybe I do need to see the QEMU forums to debug this issue.

And for the record....I'm not arguing, just trying to solve a problem, unless you want an argument, 'cause I can do both.
Next Page...
original here.