forwarding algorithm and timeouts

[Kevin Darcy]

Bob Vance wrote:

> Thanks, Kevin.
>
> >is actually assuming the role of a "resolver",
> >basically a stub-resolver-with-caching.
>
> Well, surely it's not trying the list more than once?  -- like the stub
> resolver algorithm described in 'DNS and BIND' and in contradiction to
> the documentation where it says that the "forwarders" are tried each
> only once.

Different stub resolver implementations can have different timeout/retry
strategies, as evidenced by, e.g. the disparity between nslookup's and most
system resolvers' timetout/retry strategies.

> > don't recall offhand if there
> >are any specific parameters in the RFCs regarding resolver timeouts...
>
> So, still, I'm wondering what that timeout is and the rationale for its
> value, whatever it may be.

Sorry, I don't know the specifics. I'd have to dig through the code for
that, and you can probably do that as easily as I....


- Kevin

> -------------------------------------------------
> Tks        | <mailto:BVance at sbm.com>
> BV         | <mailto:BobVance at alumni.caltech.edu>
> Sr. Technical Consultant,  SBM, A Gates/Arrow Co.
> Vox 770-623-3430           11455 Lakefield Dr.
> Fax 770-623-3429           Duluth, GA 30097-1511
> =================================================
>
> -----Original Message-----
> From: bind-users-bounce at isc.org [mailto:bind-users-bounce at isc.org]On
> Behalf Of Kevin Darcy
> Sent: Monday, March 26, 2001 5:00 PM
> To: bind-users at isc.org
> Subject: Re: forwarding algorithm and timeouts
>
> Bob Vance wrote:
>
> > What exactly is the forwarding algorithm, assuming that "forward only"
> > is set.
> > Or where can I read about it (besides the source code :) ?
> >
> > I checked the ARM, but only found this brief statement:
> >
> >    "they are queried in turn until the list is exhausted or an answer
> >     is found.
> >    "
> >
> > Hmmm. I just looked in 'DNS and BIND, 2nd' and it says:
> >
> >    "each forwarder is contacted only once, and it waits a short time
> >     for the forwarder to respond.
> >    "
>
> Actually, as of BIND 8.2.3, the forwarders are supposedly tried
> according
> to historical RTT, not necessarily sequentially...
>
> > OK.  What is the timeout used before going to the next one in the
> list?
> > I would assume that it would have to be ~5 since we know that the stub
> > resolvers wait this long on their initial try.  Otherwise, we could
> get
> > several outstanding request from resolvers before the first forwarder
> > try times out.
> >
> > Do all resolvers follow the algorithm described in 'DNS and BIND' ?
> > I.e., do WinX or MAC clients also try their list initially with 5
> > seconds, then re-do the list with 10 seconds, etc.?
> > Is this algorithm a requirement described in an RFC?
>
> In RFC terms, a nameserver doing "forward only" isn't really acting as a
> "nameserver" _per_se_, but is actually assuming the role of a
> "resolver",
> basically a stub-resolver-with-caching. I don't recall offhand if there
> are any specific parameters in the RFCs regarding resolver timeouts...
>
> - Kevin