Hi John,
Thanks for addressing my specific question. I now understand the context of your original message a lot better. To which my first response is a resounding: bit.ly
Earlier I interpreted you original post as referring to single-client provisioning, whereas you are actually referring to equalizing cable lengths by the exchange in order to present a fair playing field for 'all' clients. We have discussed the potential need for this form of arbitration, mediation in the past on this forum -- but not specifically in terms of equalizing cable lengths, per se, at any single exchange point, but more in line with establishing an algorithm-driven 'fairness' model at a higher layer of architectural abstraction. Nevertheless, I see your point more clearly now.
The point implied in my question, however, remains. By definition, equalizing cable lengths at "any" single architectural segment along the chain translates into artificially introducing a certain amount of additional latency into the paths of all players, in order to reduce their speeds to the player whose cable length is, by the happenstance of being assigned the most distant switch port, the longest.
Begin musing:
I frankly don't see how this is anything more than a half-measure, unless similar measures are taken at every hop along the routes of every client, from their most distant participating servers right up to the exchange in question, especially since, if I am not mistaken, there are multiple colos now housing those servers whose distances from the exchanges obviously differ far more significantly than the lengths of cables between racks. Otherwise I don't see this as more than a token gesture, except that it is under the singular control of the exchange, even if it does mean introducing a few picoseconds more delay to most participants. It would seem that a more elaborate handicapping system, as I noted above, would have to be created and enforced that was more reaching in a geographic sense than at the exchange switch port alone.
End musing.
On a related topic, I've recently been commenting on the low-latency arms race craze over on the Telecom Ramblings board, suggesting that we're apt to see a new category of cos/qos evolve and become integrated into vendors' wares, just as we now have for other parameters, to take latency into account in connection with SLAs, as well. This topic recently extended into the area of the growing number of Carrier Ethernet Exchanges. In response to one participant's (Optimum Lightpath) stated intentions to buy as well as sell services at Carrier Ethernet Exchanges, I wrote the following two posts, which I have only slightly edited below:
Taken from: bit.ly
The head-honcho blogger Rob Powell stated: "Optimum Lightpath, the metro fiber division of Cablevision, has now established a connection with CENX’s Carrier Ethernet Exchange. To the equation they bring some 4000 route miles of fiber connecting some 4000 lit buildings to date throughout the New York City metropolitan area. CENX boasts some 10M+ Ethernet Service Locations that can be accessed via its exchange, however I believe it is those directly connected by fiber that represent the most valuable to its development as a nexus. [...] Interestingly, ConnectedPlanet quotes Nan Chen as saying Optimum Lightpath is a buyer as much as a seller of services through the exchange. Hmmmm"
FAC says:
July 22, 2010 (2 weeks ago) at 8:35 am
Upon first reading this article yesterday it took me a moment to only begin to grasp its implications. Unless I’m all wet and totally misreading this article, the buying function places the exchange (or some other entity) in the position of a discriminating broker, taking its role beyond the rudimentary function of an Ethernet data interchange [...] unless such capabilities are automated and performed dynamically at some point on the fly. Latency, jitter, dropped data frames/packets, etc., will need to be published by participants, and made accessible either publicly or on a blind intermediary basis, which means that participants who want to have any skin in this game may be required to disclose their actual performance capabilities, which is the kind of information that is usually held very closely, which would make complying appear anathema to most operators under ordinary situations. It will be interesting to see who plays, who balks, who brokers, and the implications of all of the above. Thoughts?
--
FAC says:
July 22, 2010 (2 weeks ago) at 8:45 am
Taking the metaphor a step further, one can already begin to see how the evolution of this space will include employing protocols in intermediate nodes such as [those now used at] Layer 3, [fulfilling] the role of specialized brokers for assessing route selections based on network ‘costs’ of all types, thus further opening the door for transit and backbone players, etc., relegating the original notion of existing Carrier Ethernet Exchanges to private peering points. I second your ‘Hmmmm”.
FAC
------ |
