[300] in Resnet-Forum
Re: BootP vs Manual
daemon@ATHENA.MIT.EDU (Charley Kline)
Mon Jun 13 13:02:48 1994
Date: Mon, 13 Jun 1994 11:50:16 -0500
To: Matt Arthur <arthur@wugate.wustl.edu>, Resnet <resnet-forum@MIT.EDU>
From: cvk@uiuc.edu (Charley Kline)
At 11:02 AM 6/13/94, Matt Arthur wrote:
>As I continue down the BootP trail, the thought comes to me (via my
>director) as to why we should use BootP instead of manually assigning IP
>addresses.
>
>His contention is that the student would be able to "register" with us,
>get the IP address and then the student will be in business as soon as he
>enters it into his machine's software. That would be easier and quicker
>for the student vs. the student telling us their ethernet card address
>and us entering it into the BootP server. If a student has a Macintosh,
>how do they get ahold of their ethernet card address?
Go into the MacTCP control panel and option-click on "Ethernet".
That aside, if you manually assign addresses, what do you do if a student
helps another student out by copying his disks and installing the software,
config and all, on another machine. Then you have duplicate IP addresses
and worse, you'll have a hell of a time finding the culprit.
Basing things on hardware addresses at least allows you to keep your
database correct. Then if a duplicate IP address shows up, you know who the
"real" owner of the address is, and if your repeaters are smart enough,
what room number the offending hardware address is in.
We do this automatically at UIUC; every five minutes we poke at the ARP
table in the router with SNMP, and maintain a database of IP-to-Ethernet
address mappings. If any bogus bindings show up, a report gets generated
and the offending port is disabled automatically.
/cvk
