> Keith Moore <[EMAIL PROTECTED]> writes:
> 
> > Or you could redirect attempts to ssh to host X so that they go to
> > host Y instead.  This is not an inherently desirable feature.
> 
> Sure, but if I have write access to the DNS configuration (so that I
> can add the SRV record), then I don't need any SRV records to do that
> redirection, plain A or CNAME works fine. SRV just let's me do it on a
> more fine grained scale.

in other words, it lets you mount precision attacks :)

> > > If the server administrater wants it: Sure. Nobody else can really
> > > decide whether or not SRV is "useful" or "appropriate" for a
> > > particular service instance.
> > 
> > Nor, for that matter, can the DNS server administrator.
> 
> The server and administrator and DNS server administrator have to
> cooperate. If they for some reason refuse to do that, they simply
> can't deploy SRV in any useful way.

You missed my point, which was that the DNS server admin is usually not
in a position to know whether it's appropriate to impose SRV records on
a particular app - unless perhaps the app specifies how and when it's 
okay to do that.  SRV is not a general-purpose club that you can beat
apps with.
 
> > For instance, http://example.com:80/foo/bar will be canonicalized to
> > http://example.com/foo/bar  - which, if there were a SRV record for
> > _http._tcp.example.com pointing to a different port or host, would 
> > change the meaning of the URL.
> 
> Thanks, that's a good point. It's got more to do with URL syntax than
> with the http protocol per se,

it's pretty hard to separate HTTP and URL syntax; HTTP and URLs are very
closely tied together.

> but it still seems like it could be a
> real problem with using SRV records for http.

It's just one example among many.  Some protocols can run perfectly well
on any port; others have assumptions about port numbers wired-in , and 
often they are not obvious.  The point is that one size does not fit all,
and imposing SRV on all apps will break things.

Keith

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[EMAIL PROTECTED]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to