Re: question on key localizaton
Bob Natale <[email protected]>
| Newsgroups | gmane.ietf.snmpv3 |
|---|---|
| Message-ID | <[email protected]> |
At 11/18/2002:04:21 PM, Uri Blumenthal wrote: Hi Uri, I must say that this is a most surprising twist of events! I certainly hope that I don't just add to anyone's confusion with this input: While admittedly I can't find the e-mail trail in my personal SNMPv3 archives now, I am quite certain this issue was discussed on this list in the early days of SNMPv3 development. If I recall correctly, the explanation as to why Inform-Receivers (but not Trap-Receivers) are authoritative is that they issue the Response (or Report) PDU(s)...just like Command-Responders do...which I guess relates to the "Confirmed" PDU class thing. Personally, I preferred to consider Notification- Originators as implicit Command-Responders and, therefore, authoritative...on the basis that there should have been some prior Command-Generator like operation that directed either/both kind(s) of notifications to be sent to the respective Notification-Receiver apps. By now, of course, I have internalized the "responder=authoritative" logic (although I don't *really* consider the NR to be authoritative wrt Informs! :-). Cheers, BobN - - - - - >Wes Hardaker wrote: >> Uri> "The same key" is NOT the same as "the same LOCALIZED key". So >> Uri> yes, all those keys would be DERIVED from the user key owned by >> Uri> secName, and all those keys will be different and >> Uri> cryptographically "unrelated" due to localization process. >> >> Uri> Am I missing something here? >> >> Yes you are. The engineid for the localization process in a SNMPv3 >> inform is the notification receiver. > >If correct - it means INFORMs are badly screwed up. Depending on >how it is addressed by the net-admin, either usability goes down >the tubes, or security... > >The only correct fix that I see right now is to make the agent >authoritative engine always... Anybody remembers why we didn't >see such a simple problem and didn't address it way before? > >> Uri> One of us is missing something. The question is - who and what? (:-) >> >> It's you ;-> > >Yeah... I wish this was brought up earlier. > >Are you 100% sure it's so with authoritative engines...?