I know..I'm dating my self. ATM SVC? LOL. That was a techno flop. But... it was a great idea. Create a network that was based on switched point to point circuits. A server and a user could be on a point to point connection with eachother. What today we would call a link-local connection. It really simplifies the world.
The problem with ATM SVC was we weren't that good at network operating systems yet..and ASIC/CPU designs werent powerful enough to handle the speed and number of setup/teardown requests .
SDNs... and specifically Openflow can make this happen. We now will have a SVC like fabric...but built on ethernet. Techno flashback time... LANE was the solution for ATM.. This is better...no SAR process. Just straight ethernet..nice and efficient. No changes in technologies. You can now trick devices into thinking they are on a very very small network. One link-local to the exchange server... another to the lync server... maybe another to a nas server... maybe another to the internet firewall.
I've got thoughts on how this improves security and broadcasts as well. more about that in another blog post.
The problem with ATM SVC was we weren't that good at network operating systems yet..and ASIC/CPU designs werent powerful enough to handle the speed and number of setup/teardown requests .
SDNs... and specifically Openflow can make this happen. We now will have a SVC like fabric...but built on ethernet. Techno flashback time... LANE was the solution for ATM.. This is better...no SAR process. Just straight ethernet..nice and efficient. No changes in technologies. You can now trick devices into thinking they are on a very very small network. One link-local to the exchange server... another to the lync server... maybe another to a nas server... maybe another to the internet firewall.
I've got thoughts on how this improves security and broadcasts as well. more about that in another blog post.
No comments:
Post a Comment