Frank,
<< As for "self healing" in transoceanic rings, that's another matter entirely, and could consume many hours in discussion. Your depiction of how the "self heal" takes place in a classical sense, albeit a generic one, is accurate, and the same could and likely does hold true, from the standpoint of what is possible, for oceanic crossings as well. >>
Let's take it straight from the global crossing web site:
globalcrossing.com
"The Global Crossing Network is being engineered and constructed using the latest in fiber optic technology, including self-healing ring structures, erbium-doped fiber amplifier repeaters, wavelength division multiplexing, and the use of redundant capacity to ensure instantaneous restoration....."
I specifically referred to the GBLX network. I was not being generic at all. This is an important feature of the GBLX oceanic crossings. They use a completely redundant ring architecture that "self heals" - that's from a logical sense, not a physical one, of course. The point is, a customer buys capacity on the GBLX net with the knowledge that a single cable break/failure won't take them out of service. Before the GBLX network, (and GBLX was one of the first, if not the first, to do this on oceanic crossings - but don't quote me on that) the ocean crossings were point to point. A cable failure could take you out for for an extended time, so backup capacity had to be purchased from another system. This is very costly. If you think about it, this is a big deal.
Go to the GBLX page and look at their network map. You will notice that all of the crossings are in a "circle" or ring. Notice also that they have more than one landing site at each end. This is to protect against a failure at the landing site. If the landing site fails (building burns down, whatever) access to the crossing is available from the second site, again, automatically under the "self healing" provisions. Notice also that they interconnect a second ring to each of these end points that then traverse across a continent. In that way the complete failure of a landing site can automatically "self-heal" by switching to the redundant side of the ring and the second landing site.
Fiber rings are not new. Qwest uses them as large interconnecting rings across the United States. Rings are now used between most telco central offices to insure reliability, and most CLECs use metropolitan rings as they loop fiber between customer locations within a city.
<< Secondly, the transoceanics themselves are now boasting to be able to handle all IP over whatever, as a means of reducing dependency on SONET, if I'm reading them right. This plays into another related consideration, and that is,... >>
There is a common misperception that design decisions revolve around "sonet vs. IP". This is not a zero sum game. Sonet is a layer 1 framing standard, whereas IP is a layer 3 protocol. They both have their place (although granted wavelength division developments may make Sonet obsolete at some future time). In designing networks, you use both for redundancy and reliability where it makes sense. Rings are useful in that physical redundancy and failover occur at the physical layer. This is useful in that upper layers, ie IP, need have no knowledge of the redundant path. IP isn't even aware a failure occurred. You therefore don't have the delays in IP network convergence that routing protocols such as EIGRP or OSPF inherently have. In large networks convergence can be enough of a delay to terminate time sensitive protocol sessions such as SNA (netbios over TCP/IP), which can be a major customer issue. On the other hand, you will sometimes design failover at the IP layer because you don't want the expense of a ring architecture, or you want to protect with multiple network interfaces/hardware at the customer site. It really depends on what you are designing to.
For the GBLX network, the fact that they have a physical self-healing architecture is a marketing advantage over competitors who only have point-to-point fiber crossings.
Regards,
DN |
