Talk:Multihoming

Latest comment: 4 months ago by Jec in topic Classic, not classical

NAT-based multihoming

edit

I wrote this then did the Google search (better for the copyright :-) however, looking at this I've noticed that CISCO talks about NAT based multihoming as well. I don't think that's true multihoming, so I haven't mentioned it.

Doesn't make sense

edit

This article should explain what multihoming is in the first sentence, but I read the entire article and I still don't know what multihoming is. —Preceding unsigned comment added by 98.237.248.65 (talk) 11:32, 9 March 2011 (UTC)Reply

LDP multihoming

edit

added LDP multihoming to give readers a chance to see a different "type"/approach to multihoming

Proposed Definition to Multihoming

edit

As there is no concete definition for multihoming, I propose the following definition that is based on understanding from the following RFC document

As per the RFC A "multihomed" site is one with more than one transit provider."Site-multihoming" is the practice of arranging a site to be multihomed.

So I believe a proper defintion would be - Multihoming is a state when a site or node is connected with one or more transit providers. — Preceding unsigned comment added by Mohitkr05 (talkcontribs) 14:30, 23 January 2013 (UTC)Reply

Totally needs a re-write.

edit

Than you for writing this, but it is almost impossible to understand unless you already understand it. When the definition of a word (in this case "multihoming") contains the word itself (again, "multihoming") that's a big clear sign that something's wrong. 76.202.63.233 21:05, 26 April 2007 (UTC)Reply

What are the sources for all these ideas? Specific citations are required.

edit

You need to give credit whence these ideas came. Unless you developed this technology single handed, citations are required to the original bodies of work. It gives credit where it's due and allows the reader to look up the work of those who actually developed this technology. Not doing so is plagiarism. Wikipedia too often poses as a summary of accepted and current knowledge and wants to be respected for that. If this is true, it should make presentations in the same fashion which were developed over hundreds of years of academic inquiry.

In addition please read: Strunk, William. Elements of Style. Ithaca, N.Y.: Priv. print. [Geneva, N.Y.: Press of W.P. Humphrey], 1918; Bartleby.com, 1999. —Preceding unsigned comment added by 64.162.229.144 (talk) 16:28, 11 August 2009 (UTC)Reply

Too specific

edit

This article gives a good description of specific cases of multihoming, but multihoming is a more general concept that simply refers to multiple IPs being hosted by a single operating system. I will consider rewriting portions of this article.

Good start though... —Preceding unsigned comment added by Greatwhitesharkbear (talkcontribs) 16:32, 21 September 2010 (UTC)Reply

I would argue it is even more general than that. It is really the concept of having more than one network connection. The same concept can apply to any level and any kind of network. And "reliability" is only one reason to do it; there are many others. Best to start with the general definitions, then cite some examples. W Nowicki (talk) 21:38, 27 May 2013 (UTC)Reply

Does the final paragraph belong here?

edit

The final paragraph that starts: "Mono-homing applies if users are affiliating with a single platform. From consumers’ perspective, using ..." doesn't seem (to me) to have anything to do with the subject. If it does, it needs a rewrite. -- Dougher (talk) 17:36, 12 September 2013 (UTC)Reply

Interface vs link, what's the difference?

edit

What's the difference between an "interface" (Multiple interfaces) and a "link" (Single link, Multiple links)? The article doesn't make this clear -- interface and link aren't defined here and there are no hyperlinks to definitions. -- Dougher (talk) 18:37, 14 January 2015 (UTC)Reply

Rewrite of beginning of article

edit

I've rewritten the beginning of the article, in a way that I hope is both more correct and more clear. I'll wait a few days for feedback, then, assuming nobody disagrees, I'll rewrite the rest of the article in the same style and add some references. Jec (talk) 23:18, 5 March 2016 (UTC)Reply

I like that you've reduced the amount of jargon and clumsy language in the new version. Good work! You should work on reducing or avoiding parenthetical expressions. I'm not sure why you removed this ref: [1]. ~Kvng (talk) 02:04, 9 March 2016 (UTC)Reply
I've removed the whole section where the reference appeared ("Risks related to multihoming"), since it's not specific to multihoming, and seems non-encyclopedic to me. If you feel that the citation is relevant, please feel free to reinstate it. Jec (talk) 01:03, 12 March 2016 (UTC)Reply

Done rewriting the article. I've allowed myself to remove the multiple issues tag. --Jec (talk) 13:38, 4 May 2016 (UTC)Reply

Multihoming vs. peering

edit

I am reading the sixth edition of the Computer Networking - A Top-Down Approach, and on pages 59 to 60 it defines multi-homing and peering within the context of a simplified model of the Internet. Within a hierarchy of ISPs, where Tier 1 ISPs are at the top, lower-tier ISPs may multi-home to multiple upper-tier ISPs, which typically requires payment. Peering is different in that it is between ISPs at the same level in the hierarchy, and typically does not involve payment. Is this a common distinction? If so, maybe the introduction of this article should be made clearer. --IvarTJ (talk) 22:31, 15 June 2019 (UTC)Reply

It sounds like what is being discussed there is what's covered in the Classical multihoming section of the article. A summary of this in the lead would be welcome. There are at least two separate networking concepts being covered here. If expanded, the lead needs to make that clear. ~Kvng (talk) 01:44, 19 June 2019 (UTC)Reply

Classic, not classical

edit

I see the usage of the word classical, which refers to the Greek and Roman traditions used interchangeably with the word classic. "Classical multihoming", should be "classic multihoming". There was no network connections that I know from in the Classical Tradition. Tonskimojster (talk) 17:25, 7 March 2024 (UTC)Reply

I don't see classic or classical used in any of the current supporting sources or other sources I found in research. I'm changing classic to network. Although network multihoming also does not appear in the sources, it is at least descriptive of the application. Network multihoming is in contrast to host multihoming ~Kvng (talk) 14:14, 10 March 2024 (UTC)Reply
Network multihoming refers to multihoming at the network level, as opposed to host multihoming. I'm reverting your change. Jec (talk) 19:06, 2 June 2024 (UTC)Reply
@Jec, it looks to me from your terse reply like we're in agreement so I don't understand the revert. Can you provide a definition for classic multihoming? ~Kvng (talk) 19:24, 2 June 2024 (UTC)Reply
Classic multihoming and multihoming with multiple addresses are two techniques that can be used to implement network multihoming. While classic multihoming is the most common technique for network multihoming, it is not correct to rename classic multihoming to network multihoming, since there are other techniques for network multihoming than just classic multihoming. Jec (talk) 11:46, 23 June 2024 (UTC)Reply
I guess I'll answer my own question. I assume classical multihoming is multihoming scenario #3 in the O'Reilly source and the only type described in the Cisco source. ~Kvng (talk) 22:48, 27 June 2024 (UTC)Reply
Almost right: both (2) and (3) in the O'Reilly source describe classical multihoming. Most sources only describe classical multihoming, since it is the form of multihoming that's actively being marketed. The other forms of multihoming do not necessarily require cooperation from the service provider, and are therefore neglected in the literature. Jec (talk) 10:55, 30 June 2024 (UTC)Reply