Talk:Network topology: Difference between revisions
Jump to navigation
Jump to search
imported>Subpagination Bot m (Add {{subpages}} and remove checklist (details)) |
imported>Howard C. Berkowitz (→Emphasis on topology vs. on failure/recovery modes: new section) |
||
| Line 4: | Line 4: | ||
:Ugh I just need more time lol! --[[User:Eric M Gearhart|Eric M Gearhart]] 12:02, 8 April 2007 (CDT) | :Ugh I just need more time lol! --[[User:Eric M Gearhart|Eric M Gearhart]] 12:02, 8 April 2007 (CDT) | ||
== Emphasis on topology vs. on failure/recovery modes == | |||
There are a number of topologies that apply when high availability is involved, not just in physical media such as SONET/SDH/NGN, but in things like the MPLS recovery framework (RFC3469 Framework for Multi-Protocol Label Switching (MPLS)-based Recovery. V. Sharma, Ed., F. Hellstrand, Ed.. February 2003) | |||
Some things to consider, with fault tolerant rings, are: | |||
:*1+1 data flows concurrently on both rings, and the endpoint decides which to use | |||
:*1:1 data flows on an active ring, but can fail over to a protection ring. The failover, incidentally, can be for less than the entire ring--a failing station or media segment can be bypassed | |||
:*N:M there are N active rings protected by M backup rings, such that N>M | |||
[[User:Howard C. Berkowitz|Howard C. Berkowitz]] 12:39, 12 May 2008 (CDT) | |||
Revision as of 11:39, 12 May 2008
Need some ideas for more topologies? Think wireless: broadcast, multi-hop. Markus Baumeister 15:17, 2 April 2007 (CDT)
- Ugh I just need more time lol! --Eric M Gearhart 12:02, 8 April 2007 (CDT)
Emphasis on topology vs. on failure/recovery modes
There are a number of topologies that apply when high availability is involved, not just in physical media such as SONET/SDH/NGN, but in things like the MPLS recovery framework (RFC3469 Framework for Multi-Protocol Label Switching (MPLS)-based Recovery. V. Sharma, Ed., F. Hellstrand, Ed.. February 2003)
Some things to consider, with fault tolerant rings, are:
- 1+1 data flows concurrently on both rings, and the endpoint decides which to use
- 1:1 data flows on an active ring, but can fail over to a protection ring. The failover, incidentally, can be for less than the entire ring--a failing station or media segment can be bypassed
- N:M there are N active rings protected by M backup rings, such that N>M
Howard C. Berkowitz 12:39, 12 May 2008 (CDT)