Barbara, Humble apologies for the URL to the Cisco ISIS diags document. Please use the tinyurl below which works. I will read rest of your email later in the day and other emails from homenet as well.
http://tinyurl.com/o8znoam Hemant -----Original Message----- From: STARK, BARBARA H [mailto:bs7...@att.com] Sent: Tuesday, July 28, 2015 10:24 AM To: Hemant Singh (shemant); homenet@ietf.org Subject: RE: some IS-IS questions Hi Hemant, Thanks for the reply, but... > >There was a claim that IS-IS provides "diagnostics". > >What sort of diagnostics? > > http://www.cisco.com/c/en/us/td/docs/routers/asr9000/software/asr9k_r4 > 3/routing/configuration/guide/b_routing_cg43xasr9k/b_routing_cg43xasr9 > k > _chapter_01011.html#concept_B475330101904631B14C625066A033C3 > > Search at the above URL for " Route Convergence Monitoring and > Diagnostics". The link didn't work for me (and, yes, I made sure the entire string was there), but I searched on the Cisco site for ASR 9000 documentation and found instructions for configuring/enabling this Cisco function, and a statement that such a function exists. I still have no clue what this function does -- what the detailed output is, and how to interpret. Some people say they want IS-IS because it can also do some sort of diagnostics, but I still have no idea what these diagnostics are. BTW, I did do a quick price scan of Cisco ASR 9000 series routers, and believe they may be just a little bit outside the range of the average consumer. So I don't know if it would be reasonable to find this RCMD functionality in a consumer-grade router. > >Is this a reference to topology discovery? Does topology discovery > >rely on > device participation in IS-IS, or can devices that do not participate > in IS-IS be discovered? > > All devices in the IS-IS domain need to be routers. Neighboring routers find > each other by exchanging Hellos using multicast. Loosely speaking, an IS-IS > router uses link-state packet (LSP) to flood learnt prefixes and > topology to neighbors. IS-IS supports both IPv4 and IPv6 easily. So when IS-IS talks about topology discovery, it's talking about router topology, with no knowledge of hosts or bridges or PHY technologies. I'm sorry, but in a home network, the router topology is really the least of my worries. > >Can bridged devices be discovered? > > Yes/No. The bridge would need to be a RBridge such as in TRILL and then the > bridge can be discovered. Or if the bridge is connected to an IS-IS router > or a > switch connected to the IS-IS router, the router/switch can learn the > end bridged device mac-address. The learnt mac-address can be > propagated to IS-IS using a sub TLV in IS-IS. So the answer is no. Bridges are not going to be discovered in a homenet IS-IS topology. > >Can physical layer topology be discovered? > > IS-IS knows who its neighbors are and also knows the path to any other > neighbor. Wouldn't this suffice? Or please give an example of a physical > layer topology that the network layer is not able to provide. No, this does not suffice. I'll have to spend some time putting together my physical layer examples. In short, the vast majority of problems in the home network are due to issues of physical connectivity -- links go up and down causing consumer electronics devices to "lose" their IP address(es) and stop working until something kicks the device's network interface to get an IP address again (with reboot being the easiest and most-likely-to-work form of kicking). In short, I'm slowly coming to the following conclusions: 1. Whatever diagnostics / topology discovery mechanisms may exists in IS-IS are insufficient to be of any real use. So any argument for IS-IS based on the existence of such diagnostics is irrelevant. [But I can be swayed from this conclusion if someone can provide real info showing this conclusion is wrong -- in an IS-IS load suitable for homenet routers.] 2. Technologies that are not resilient against links that go up and down frequently and for no apparent reason are useless in a home network. These links are prevalent in the home network. And not just the wireless links. The powerline and even coax links are also subject to this problem. In my experience, these up-and-down links are The Number One Cause of home networking issues today. Barbara _______________________________________________ homenet mailing list homenet@ietf.org https://www.ietf.org/mailman/listinfo/homenet
