What’s the best way to modify network interfaces from the command line?
Basically, I’ve got a cluster node set up as a router in one building (mainly offsite backup, but it’s also routing traffic to/from an AirPort Base Station to extend a branch of our network) and it has no video card (nor do I have access to displays at that location even if I put a video card in it).
As the server is off site, I did the install and configuration in a lab environment and then just popped the drive in the server, booted it up, and thing seemed to run fine. It’s running NAT, DHCP, and such.
However, at one point I needed to make a minor change to the configuration for the two ethernet ports (and I’ve got to add another one in at some point soon, as well)… something basic, probably adding/modifying DNS entries or something… I did so via ARD and at that point System Preferences seemed to have detected that the ethernet ports had new MAC addresses, so it duplicated all the ports. this made it a nightmare for getting the settings right and basically broke NAT.
Unfortunately, it was an easier solution to re-image the drive in my lab environment and pop it into the server at the time than anything else. However, I’ve got more changes I’m going to need to make in the near future and would like to figure out how to make corrections to the network interfaces via the command line (over the serial port) if I mess it up again.
I haven’t done such configuration since I was running NetBSD 1.5 on some older Mac hardware, so my skills in that area are basically non-existant or atleast inappropriate.
Can I just use ifconfig? If so, will changes made be preserved over a reboot? Will they show up correctly in System Preferences? Any pointers at the syntax?
And last, but not least, how does one make sure that the interfaces don’t get duplicated like that? Do I have to pull the Xserve out of that rack and actually to the install directly on it?
.
.
Comments are closed