Re: [ntpwg] Different security mechanisms (NTS, Autokey, Symmetric Security) and key exchange techniques

[Harlan Stenn <[e-mail]>]

Hal Murray writes:
> 
> [e-mail] said:
> > Specifically, how much time is spent processing each of the first 6 packets
> ?
> >  I'm curious how many of these steps are expected to take 1.75 seconds or
> > longer. 
> 
> Where did the magic number of 1.75 come from?

It's less than 2, and leaves some room before it becomes 2.

iburst packets are sent at 2 second intervals.

H
--
> There is a big difference between "expected" for the normal case and 
> something like "happens often enough" so we shouldn't rule it out.
> 
> Are you asking about CPU time for client or server to process a packet or are
>  
> you including network transit time?  In either case, it's not hard to come up
>  
> with nasty cases that take a long time.
> 
> Bufferbloat can easily get over 2 seconds.  My DSL line is close to 4.
> 
> I leave it to your imagination as to how a CPU can take that long.  My straw 
> man would be a busy system without enough memory so it's swapping.  Logging 
> to a stuck NFS server is another good choice.
> 
> 
> -- 
> These are my opinions.  I hate spam.
> 
> 
> 
> 
_______________________________________________
ntpwg mailing list
[e-mail]
http://lists.ntp.org/listinfo/ntpwg