We've been through this before. History does repeat itself...
You have a simple protocol, like Ethernet, but someone comes along and say, naw...we can make it better, lets take a protocol that is well known, works, and is simple and lets make it not only overly complicated..but we'll make it almost impossible to be inter-operable. You want broadcast and multicast? Yeah..we can give you that..but it probably wont work most of the time..because the underlying technology is inherently point to point.
If you grew up in your networking career in the 2000's.. you probably missed that one. Thats ok... we're in the middle of the next overly complex technology, FCOE. Yes, take a very simple, proven technology that is deployed all of the place. Well known. Then stick it on top of another technology that doesn't inherently support some of the basics things the other one does. Namely... being lossless, consistent latency, loopfree, and the ability to multipath.
Sure, we there ways we can get around these issues..but low and behold... they are not inter-operable.
The problem with FCOE as well, as it doesn't inherently change anything. It doesn't give me more bandwidth, it doesn't speed up my storage, and it doesn't help me with the cloud or with multiple data centers.
Dont mess with this decade's LANE..it will leave you burned.
You have a simple protocol, like Ethernet, but someone comes along and say, naw...we can make it better, lets take a protocol that is well known, works, and is simple and lets make it not only overly complicated..but we'll make it almost impossible to be inter-operable. You want broadcast and multicast? Yeah..we can give you that..but it probably wont work most of the time..because the underlying technology is inherently point to point.
If you grew up in your networking career in the 2000's.. you probably missed that one. Thats ok... we're in the middle of the next overly complex technology, FCOE. Yes, take a very simple, proven technology that is deployed all of the place. Well known. Then stick it on top of another technology that doesn't inherently support some of the basics things the other one does. Namely... being lossless, consistent latency, loopfree, and the ability to multipath.
Sure, we there ways we can get around these issues..but low and behold... they are not inter-operable.
The problem with FCOE as well, as it doesn't inherently change anything. It doesn't give me more bandwidth, it doesn't speed up my storage, and it doesn't help me with the cloud or with multiple data centers.
Dont mess with this decade's LANE..it will leave you burned.
No comments:
Post a Comment