In my mind, a suite of practices to keep one's garbage contained and not all over the neighbor's lawn is a good thing and covers many bases. RPF/BCP38 seems to be the IP level equivalent of blocking ingress SMTP and forcing delivery through outbound-only servers that check the claimed envelope and/or header senders for sanity relative to the authorized sending networks. If so many people are agreeing on BCP38, what's with the resistance about email, clearly an equally polluted swamp? Why would one not want to view the two issues as much the same problem, at different layers? And yes, I was assuming split-brained mail infrastructure to make port-25 filtering much simpler. To counter someone's counterargument, it could boil down to two ACL lines in *many* places, but clearly not all. Said two lines can come right before the one that says "permit ip my-source-only any", couldn't they?? Not in a blanket sense, of course -- these things done *where appropriate* and tuned to known requirements could vastly improve matters, but it seems that even after all these years so many of the appropriate places haven't even been touched let alone fixed. _H*
Well at the risk of getting flammed here.. lol I don't believe there is a real clear answer here to this BCP38 debate. Great suggestions, great comments, and great what ifs.
From the old days, I always recalled ACLing non-existant scopes within my nets, again not that that is the answer, but it was a recommended practice, and when we saw non-existant spaces trying to leave one of our feeds it was quickly handled internally (i.e. killed the downstream link). As well we always had to do an internal audit of why/who/where the event took place and a remedy to it (HIPAA & SOX compliance stuff)
While this thread is informative at times, I think the name calling and insults really serve no purpose to it. I recall a funny saying regarding this, opinions are like a......s, everyone has one and everyone else thinks it stinks. Doesn't mean anyones right. Agree to dis-agree and lets be on with it. Deja-vu, Wasn't there a thread about this same subject a while ago something regarding RFC2827? Might just be me. Just my 2¢s Regards, -Joe Blanchard "I am Joe Blanchard and I approve this message.... lol"
participants (2)
-
hobbit@avian.org
-
Joe Blanchard