So what’s the risk with SDN?

We deceive ourselves when we fancy that only weakness needs support. Strength needs it far more.”
Anne Sophie Swetchine
.

We’ve been extolling the virtues of SDN for some time now. But how about possible weaknesses?

In an earlier post entitled Carrier SDN we discussed the potential reduction in time to market (TTM) in a software definable network. Innovation at the speed of software as it were.

However, I fear this adaptability could actually be a double-edged sword.

In another earlier post entitled Ruthless Simplification we discussed the need to keep things simple, the need to reduce the number of variables that the B/OSS needs to deal with.

The current network approach sees CSPs dependent upon their standards bodies and vendors to make decisions relating to new services or features on behalf of the whole industry. This slows down TTM, but it also significantly reduces change and variability, not to mention the extra rigour that generally goes into each new incremental change.

Are we creating a Frankenstein of unlimited change? I look forward to seeing what mechanisms the CSPs need to put in place to put controls around this variability. It’s strength also needs support.

Read the Passionate About OSS Blog for more or Subscribe to the Passionate About OSS Blog by Email

2 thoughts on “So what’s the risk with SDN?

  1. Well said, and I’d perhaps add that SDN, generally pursued with NFV, for all its benefits presents a really major risk also early-on in its existence, in that it raises the requirement to perform fault prediction and analysis in a new multilayer environment with far more variables than in current networks. The potential for a significantly larger number of suppliers in the typical SDN/NFV enabled network and particularly the vertical layering of their offerings, with more inter-supplier interfaces required, presents new levels of complexity to be reflected in analysis tools and requiring experience. The number of test cases is likely to rise exponentially along with the number of possible combinations, vs the “same full function box deployed worldwide” model prevalent today.
    Best Frank

  2. Hi Frank,

    So true! I guess we will all be waiting patiently to see how the vendor and standards “wars” plays out and what level of interoperability testing will be necessary for CSPs across the globe. As a case in point, I have no doubt that software defined networking of some form will play a big part in the future of comms, but it remains to be seen whether it’s the SDN that has evolved from work done by Stanford and Berkeley through to ONF’s leadership. The guys in the network and OSS testing departments must be really looking forward to this stuff getting out into the wild! 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *