Network connection from several linux machines connected to Whonix gateway suddenly broke

Support
1

I have a few Debian Jessie vm’s connected to the Whonix gateway, but after having cloned one of the vms the network connections from all the Debian Jessie machines to the whonix gateway broke. I can remember I checked reinitialize the mac address of all network cards, maybe this does have something to do with it.

This is my /etc/network/interfaces in my Debian machines, network manager is uninstalled, used to work perfectly:

auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
address 10.152.152.15
netmask 255.255.192.0
gateway 10.152.152.10

The Debian Jessie VM has 1 adapter: PCnet-FAST ||| (Internal Network, 'Whonix')

This is my /etc/network/interfaces (without comments) in the Whonix Gateway:

auto lo
iface lo inet loopback

auto eth0
iface eth0 inet dhcp

auto eth1
iface eth1 inet static
address 10.152.152.10
netmask 255.255.192.0

Whonix gateway has 2 adapters:

  • PCnet-FAST ||| (NAT)
  • PCnet-FAST ||| (Internal Network, 'Whonix')

Strange thing is, the Whonix workstation that is also connected to the same gateway continues to work, while several linux machines that used to work in the past don’t anymore.

What else is the Whonix Gateway doing to make internal network connections possible? What options do I have left to troubleshoot this problem? Or have I maybe uncovered a bug in Virtualbox?

2

Good day,

Does this also apply when only running one custom Workstation alone?

Have a nice day,

Ego

1 Like
3

Is /etc/network/interfaces identical in all of your Debian VMs?

Besides having unique MAC addresses which you ensured, each VM needs to have a unique IP address.

The final quartet needs to be different in each machine. Anything from 11 to 255 will do.

1 Like
4

Strange enough no. Workstation keeps working. That’s why I’m posting maybe there’s something Whonix does that is different then the jessie vm’s.

5

They are not identical, the mac addresses are also all different. However, upon checking this, one of the vm’s network is still functioning. I think I hit a bug which probably takes too much time to figure out. I’m going to recreate the other ones with the one that’s still working.

Thanks for your help guys!

6

Good day,

I was asking about whether running one CUSTOM workstation alone works. Would you mind trying?

Have a nice day,

Ego

7

My case was caused by allocating too less memory (128M), which cuased gateway killing Tor process.
Looks like when more workstations use gateway it need more memory.
256M solved the issue.

1 Like