Hi, Thanks a lot for your comments (again).
>De: Pekka Savola [[EMAIL PROTECTED]] > >I'm also bringing this to the w.g. list, as I believe an RA option for >discovering DNS resolvers could be one very viable mechanism. > >Generic comments: > > 1) It's not clear to me whether this option is to be included in periodic >unsolicited router advertisements, or only in explicit request-reply >pairs. Based on the fact that the option includes Code reply/query pair, >this seems likely to fall somewhere between the two of these. > >The way spec is written, the routers which would support DNSR option would >not include it in every RA message. (This could be considered both a good >and a bad thing.) > LB: My idea was to reduce the use of that option at a maximum. That means that routers would not have to send DNSR option in unsolicited Router Advertisements. But a node would have to ask for that option with a DNSR Query option in a Router Solicitation. Moreover the same node should also listen to DNSR option so as to not ask for that option if another node as just ask for one. What do the wg think about ? > 2) perhaps it should be explicitly stated, for clarity, that unrecognized >DNSR options must be ignored by receivers by RFC2461. > LB: ok it can be clarified > 3) Security considerations could use some additional considerations >related to adding DNS resolvers in the RS/RA messaging. At least it >should be mentioned -- I don't think there's necessarily all that >dangerous stuff out there, as if you use RA, you already have quite a lot >of trust in the system. > LB: I'm not an expert in security issues, could you explain a bit more? I have some ideas but I'm actually not sure. > 4) On systems which configure manually ie. disable RA-advertisements, I >believe this will also not be used? > >Specifics: > >The purpose is > also to enhance the efficiency of [DISCOVERY] when DHCPv6 servers can > not be used. > >==> s/servers can not be/is not/ > >4.2.2 Passive mode > > On any link a node MUST listen to Router Advertisements (DISCOVERY]. > >==> MUST is news to me. Did I miss something? > LB:... I have just read again RFC2461. Yes the "MUST" is incorrect. LB: Thanks for other comments, I will add those changes is the next version. Luc -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
