Video: Being evil with fibre in the lab

av Fredrik "Hugge" Korsbäck den 10 Nov 2015

We are doing intensive testing with the hardware and software for the new SUNET. Last week I wrote about the acceptance-test of the new routers and how it was setup and seemingly everything is good so far. This week we have installed the new optical line-system with all key-components such as amplifiers, roadms, filters, muxes and such.

There is a more thorough blogpost coming on the optical network soon but I wanted to put this out as a little teaser. On the very detailed test-protocol we have developed there is a lot of different scenarios we need to test out and see how the gear is performing and that all vendors is living up to said promises in the tenders.

One scenario we dreamed up was that we setup a regular optical span with a coherent 100Gbps signal originating directly on the routers, injecting this onto about 90km of fibre including EDFA amplifiers, multi-degree ROADMS and then somewhere in the middle of this we start banging on the fibre with a hammer. Also we run a full 100Gbps client-signal through this network.

We wanted mainly to test on how the optics would perform and how the FEC-calculations would shift when very sudden changes in attenuation and SNR would be introduced into the network. The fibre we tested this on is regular G.652.



The testprogram will show green in all boxes if no measurable error is noted on the link, if it swaps to red its unrecoverable errors that will generate bit-faults.


Fredrik "Hugge" Korsbäck

Network architect and chaosmonkey for AS1653 and AS2603. Fluent in BGP