An update on the optical build
It's been a while since I've posted, so I wanted to clear the air a bit and do a big update on where we stand with the network build to date.
First, we've been clear and done with Phase 1 since last fall. That part of the network is carrying live traffic and is completely in production. The last several months have focused on installation and turnup of the Phase 2 optical build. There, I'm happy to report, we're almost done! The map below details progress to date.
Phase 3 should be right behind Phase 2. We placed orders for all of the Ciena optical equipment in mid January. Some of it has begun to trickle into our staging facility in Broomfield, CO. A few of us had an opportunity to tour the warehouse that receives all the equipment at Level(3) a few weeks back.
Internet2, ESNet, Ciena and Level(3) representatives at the Level(3) staging facility in Broomfield, CO (Photograph supplied by Internet2 Vice President for Network Services, Rob Vietzke)
We're expecting the remaining equipment to arrive throughout March, enabling the build to proceed without pause. We are going to be performing a system upgrade in the coming weeks that will allow support for some newer transceiver cards and some other management niceties on the Ciena platform.
In terms of network transition, we're starting to formulate plans to migrate networks over to the new Ciena infrastructure in the coming months. I spent most of 2007 working with Internet2 Connectors on migration from the Qwest footprint to the new Level(3) POP. This migration in 2011 will be a cakewalk compared to that. Just a few fiber swings on the remote ends of a circuit and the work will be done.
We're spending most of our time working on the remaining phase 3 colocation and modeling our transceiver slotting. We have 118 40G muxponders working their way to Bloomington, Indiana. Each pair of those cards will provide four 10G waves between optical nodes. We'll need these to perform the migrations I mentioned above. But we want to slot them in the best places possible. We also want to use them to do a bit of optical reach testing. This will tell us how close the software-based network modelers are to the real world in terms of how far we can push a circuit without regeneration. It should be a busy few months of card slotting and network configuration, starting in late March.
Finally, I want to acknowledge an area where the network is beginning to produce some real-world change for the Internet2 community. My colleague Randy Brogle will be posting soon about some new 100G connections to the network that we're very excited to start providing as the network becomes reality. Look for his post soon!