- From: Dave Dolson <ddolson@sandvine.com>
- Date: Tue, 4 Nov 2014 18:53:59 +0000
- To: Amos Jeffries <squid3@treenet.co.nz>, "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
I'm just seeing this at a high level, so I'll just throw this idea out there: Don't bother with DNS names; you could use a well-known multicast IP address, similar to other discovery protocols. E.g., IPv6 multicast address FF05::1:3 is used to discover DHCP agents. Or use a dedicated anycast IP address. E.g., 192.0.0.1 used in RFC 6333. -Dave -----Original Message----- From: Amos Jeffries [mailto:squid3@treenet.co.nz] Sent: Tuesday, November 04, 2014 1:21 PM To: ietf-http-wg@w3.org Subject: Re: WPD Discovery Options -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 5/11/2014 4:55 a.m., Martin Nilsson wrote: > On Mon, 03 Nov 2014 14:58:49 +0100, DIEGO LOPEZ GARCIA > <diego.r.lopez@telefonica.com> wrote: > > > > WPD discusses discovery of proxies using a well known URI. > Automatic proxy discovery may be desirable by some. The choice of > an automatic proxy discovery method should be decided by the > client. Below is a selection of possible methods: > > mDNS / DNS-SD: the best all round solution. DNS can query a > well-known/predefined WPD hostname(s) and is currently supported by > wide variety of clients. One drawback is that DNS can be spoofed so > would require a trusted network. > > PDP/PDN Context: best for working with a mobile network. It is also > possible to use the PDP/PDN signalling also over WiFI etc for > configuring parameters, though in WiFi scenarios tethering can > become an issue: a client could detect the network type to the next > hop which could be a tethered wifi connection but onwards from this > point could be any type of network. (Any suggested solutions to > this are welcome!) > > > I think it is important to find a single method that works > everywhere, if possible. Once that is in place, it would be > possible for other actors to optimize that for specific domains. > E.g. if we standardize on a DNS based discovery, you could push the > relevant records to the device in PDP context. Higher layer code > still only needs to know about a single method. At first look SRV records under the .local domain should be serviceable in all networking situations. As you say the records or results can be optimized by pushing via any other means appropriate to the specific use-case. AYJ -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (MingW32) iQEcBAEBAgAGBQJUWRkGAAoJELJo5wb/XPRj+vEIAIENhktWo5WuPjnUBidUAqs9 j53iamKdNl3GNRQ8eJXAFCynoaoEwchw8I8qaWzQ+fpsgVDkRmtNoVaAoF0lDrbr V7ly8zSxHaXmuiuEXbIUll/cHF0fwbN/QTNTOqQRyd0Jwxb7hSWZK2h86gho9p6G Xciil5uCFy1q6OptLtuF7exbxssGc4jgwtRdUNykPE6bQ1dcgudG58W8y/LLw5sf qySl/OpG5CZTgbg5Zz4zoHsyWSRz5Wu+9EgiMPPCT50J8+6Xs9yPHpdUo0v/sJDf O0kS1ZNGLyq/FgLUG+DzmMu3L+8QwCkORU0YI7CXiGEiJNe2pd231rE0HlF8Kqg= =l7ky -----END PGP SIGNATURE-----
Received on Tuesday, 4 November 2014 18:54:21 UTC