Hi Martin, Thank you for your comments. See in line.
> -----Original Message----- > From: [email protected] [mailto:[email protected]] On Behalf Of > Martin Stiemerling > Sent: Tuesday, July 13, 2010 5:39 PM > To: [email protected] > Subject: [alto] Review of draft-song-alto-server-discovery-03.txt > > Hi, > > Here is my brief review of draft-song-alto-server-discovery-03.txt: > > The draft discussed various approaches to the discovery of ALTO servers, > such as DNS-based, DHCP, IANA databases, etc, but lacks a technical > discussions of pros and cons of each approach. > The approach that will be used by specific ALTO service provider will be based on its deployment scenario. > For instance, DHCP options are good to have, but it needs all elements to > understand the new option (DHCP servers, DHCP clients, DHCP proxies > (homegateways -- yeah!). Second, a DHCP option is limited to discover only > ALTO servers that are provided by the ISP but not beyond (which is not in line > with the ALTO requirements). > The DHCP proxies are not required to support the new DHCP option. See second paragraph in section 5.1.2.1. I hope the DHCP Container Option could be helpful to solve all the similar problems. Again, the discovery approach used by specific ALTO service provider will be based on its specific deployment scenario. An ALTO service provider who provides the service to a campus and another one who provides the service to a whole city may use different discovery approaches. > IANA database: This is a no go, as the IANA provides registries for code > points but not for services. > The database or WHOIS service can be used to determine the domain name, which is a step for the service discovery. We also said that it is not a good approach to do it in the draft. The reasons can be found in section 5.1.2.2. I don't want to reply to the following items one by one. They basically said the same question. Unless we can have a general discovery mechanism suitable for all scenarios. Otherwise, it is deployment-scenario-dependent. I will be very happy to see if there is a discovery mechanism without any limitation:) Thanks again, Haibin > XRD: This probably goes beyond the type of service discover we need, as the > service provided boils down to an IP address and port number to be looked > up. > > Provisioning: This not an option, as how do you distribute this file and > manage it (revoke it, update it, etc) > > Manual configuration: We shouldn't even consider this at all: What happens > if you manually configure the ALTO server and you move with your laptop to > another coffee shop, that has a different WLAN connected to a different > ISP? > > Multicast: I have no idea why people do always come back to multicast. > There is no multicast Internet-wide. > > The draft has a tutorial style (i.e., a list of possible discovery mechanisms), > but lacks IMHO the real discussion about the various options and a protocol > specification (or at least a light start of it). > > Long story short: I miss the meat in the draft. > > With kind regards > > Martin > > [email protected] > > NEC Laboratories Europe - Network Research Division > NEC Europe Limited | Registered Office: NEC House, 1 Victoria Road, > London W3 6BL | Registered in England 2832014 > > > _______________________________________________ > alto mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/alto _______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
