[ntp:hackers] [ntp:questions] Question on findexistingpeer

David L. Mills mills at udel.edu
Sun May 20 08:34:13 PDT 2007


Rajaram.R,

 From a protocol point of view there is no reason to forbid more than 
one mode association with the same server. It's not very useful, but not 
an error.

Dave

Rajaram wrote:

> David
>
> Thanks for the response. To my understanding a NTP device should not 
> have more than one mode of peer connection with another NTP device. Is 
> my understanding correct ? As a hacker/user i would like to know what 
> would be the scenario to cover such code or it is a dead code.
>
> TIA
> Rajaram.R
>
>
> "David L. Mills" <mills at udel.edu> wrote in message 
> news:f2neao$jdv$1 at scrotar.nss.udel.edu...
>
>> Rajaram,
>>
>> If the shoe fits, wear it. Do you have a problem with that?
>>
>> Rajaram wrote:
>>
>>> Hi experts
>>>
>>> I have a very basic question. The reference implementation has an 
>>> api findexistingpeer. The API returns the peer structure for the 
>>> given address. It also checks the mode before returning.
>>>
>>> -snip-
>>> if (mode == -1)
>>> return (peer);
>>> else if (peer->hmode == mode)
>>> break;
>>> -snip-
>>>
>>> My question is that whether a peer(address) can associate with 
>>> another peer with more than one mode? If yes what would be the 
>>> scenario ? If no, why findexistingpeer checks mode ? Please 
>>> enlighten me if i miss something ?
>>>
>>>
>>> TIA
>>> Rajaram.R
>>> _______________________________________________
>>> questions mailing list
>>> questions at support.ntp.org
>>> https://support.ntp.org/mailman/listinfo/questions
>>>
>> _______________________________________________
>> questions mailing list
>> questions at support.ntp.org
>> https://support.ntp.org/mailman/listinfo/questions
>
>



More information about the hackers mailing list