Hi Ron, OK. I understand your point.
Best regards, JMC. 2017-11-06 19:49 GMT+01:00 Ron Bonica <[email protected]>: > Jean-Michel, > > > > I understand that Probe addresses two operational needs. One need is to > probe the local node and the other is to probe a directly connected node. > But in either case, Probe semantics must be consistent. > > > > Probe queries the status of exactly one interface. Regardless of whether > the L-bit is set or clear, if two interface satisfy the Extended Echo > Request Match criteria, the Extended Echo Reply must return an error code > (3). Returning any information beyond that would be risky. For example, > assume that both interfaces are active, one is running IPv4 only and the > other is running IPv6 only? What would the Extended Echo Reply return? > > > > > Ron > > > > > > *From:* Jean-Michel Combes [mailto:[email protected]] > *Sent:* Monday, November 6, 2017 7:09 AM > *To:* Ron Bonica <[email protected]> > *Cc:* [email protected] > *Subject:* Re: [Int-area] Intdir early review of > draft-ietf-intarea-probe-00 > > > > Hi Ron, > > > > 2017-10-31 17:45 GMT+01:00 Ron Bonica <[email protected]>: > > Hi Jean-Michel, > > > > The semantic of this utility is to probe exactly one interface at a time. > If multiple interfaces satisfy query’s match conditions, Probe can’t > provide a response. > > > > I agree that Probe could set the A bit if all of the interfaces are > active. But what should it do if one interface is active and the other is > not? > > > > <JMC> > > > > Your opinion differs from my opinion because, IMHO, yours is mainly based > on the "probed interface resides on the probed node" (i.e., it is possible > to identify some interfaces are active and some not) and mine is mainly > based on the "probed interface doesn't reside on the probed node" (i.e., > ARP/NC allows only to identify active interfaces). > > Correct? > > Best regards, > > JMC. > > > > </JMC> > > > > > > > Ron > > > > > > > <JMC> > > Question: > > Why the A-bit is not set when Code Field is equal to Multiple Interfaces > > Satisfy Query (3) and the L-bit is clear? Question: Same question when > L-bit is > > set. </JMC> > > > > <snip> > > > [RB ] > > Error code 3 (Multiple Interfaces Satisfy Query) means that they query is > ambiguous. Or, on other words, that two or more interfaces satisfy the > match conditions specified by the query. So, we can't set the A, F, S or E > bits because we don't know which interface the user is asking about. > > > > <JMC> > > But, even if this is ambiguous, IMHO, all the interfaces are active => > A-bit should be set > > And Error Code 3 returned to probing interface raises this is ambiguous. > Now, I am OK that F, S and E should be cleared to avoid any confusion > between the different interfaces. > > > > Did I miss something? > > </JMC> > > <snip> > > > > Best regards, > > JMC. > > >
_______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
