Dear Gustavo, > > I have some comments on the mechanism described in song-discovery > for the ALTO server discovery by end terminals: > > It seems to require the configuration of the ALTO DNS SRV name in > the DHCP server. Perhaps this could make more complicated the > deployment of the solution. It could be possible to use a reverse > DNS lookup to discover the ALTO service name. This was > previously described in BEP-22.
IMHO, in this scenario, DHCP is an simple way to get the ALTO service name although you need the configuration on the DHCP server in ISP network. If you use BEP-22, there is still a lot of work to do, e.g., you need to implement STUN servers or other mechanisms to find your public address, and the hierachical network address structure makes it a little more complicated when you don't depend on some central mechanism. I am also a little conf used with using the DNS reverse lookup for host name when I have a host name xx.xx.myCompany.com and the ALTO service is provided by my ISP with the service name _alto._tcp._myISP.com. > The SRV based DNS discovery allows to retrieve the IP Address and > port of the ALTO server. A mechanism based in U-NAPTR could be > more flexible and allow the discovery of a url (for example for > the case of InfoExport where a HTTP url is needed). This U- > NAPTR based discovery is used in other discovery protocols like > draft-ietf-geopriv-lis-discovery-07. Thanks for the good reference. I will try to understand that draft. Anyway, the existing DNS SRV mechanism works. Best Regards! Song Haibin > Best Regards, > G. > > ________________________________________ > From: [email protected] [[email protected]] On Behalf Of > songhaibin 64081 [[email protected]] > Sent: Thursday, March 05, 2009 10:25 AM > To: [email protected] > Subject: [alto] ALTO server discovery > > Hi all, > > We also submit an ALTO server discovery draft. Here is the link: > > http://tools.ietf.org/id/draft-song-alto-server-discovery-00.txt > > Now there are two drafts on ALTO server discovery. I think they > have commonness and differences. Our draft mainly focuses on the > scenario where network operators provide the alto servers. The > solution has the consideration for residential gateways when using > DHCP option to retrieve the alto service name. We also take into > account the discovery process by end terminals and application > trackers(or proxies). These two drafts can be of compleme > ntary to each other. > > Any comments are appreciated! > > Best Regards! > Song Haibin > _______________________________________________ > alto mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/alto > _______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
