Open podfeet opened 4 years ago
I've researched this extensively and leaned on a couple of smart network buddies of mine who used to work for me. We've concluded that his explanation is definitely incorrect.
The Magic Cookie was at one time an option called "vend" that the vendor could choose a value (see RFC951 RFC951 but in 1988, RFC1048 changed it to the fixed value 63.82.53.63 (or 0x63825363 as tcpdump returns it).
From reading forum posts, I believe this is a correct statement:
The presence of the Magic Cookie tells the server to return the DHCP protocol, not BOOTP.
Debating how to deal with this. I could leave it wrong in the book, tell him after we give him the book and let him fix it. Or I could fix it now and then confess after the fact. Or I could try to remove all references to it. I'm leaning towards leaving it wrong but it might kill me.
Let's leave it for now and come back to it later. It's easier to change our minds later when the original text is still there. I guess we will have more issues to go back to in hindsight.
Since we're close to publishing it and handing it over to Bart, I suppose we should leave it up to him to correct. All the information is here in the issue that helps write the correct version. After rereading it, the magic cookie is still part of the DHCP packets, so the reference to it, is ok. The only lines that need to be changed are 124 and 125.
Agreed - so do we close the issue even though we need to get back to it?
Let’s leave it open for easier reference. No harm in publishing with still a few issues open.
In this chapter Bart says the magic cookie that is revealed using
tcpdump
is a randomly generated value and is used to uniquely identify the device requesting the IP. I'm pretty sure he's wrong, as my computer is generating the same exact magic cookie value as in his example. I also checked the Wikipedia page on DHCP and it's showing the same magic cookie value. I'm researching and asking Knightwise's advice.