Russian Submarines and the battle between Cloud and Edge Computing

Russian Submarines and the battle between Cloud and Edge Computing

Today 99 percent of transcontinental Internet traffic depends upon cables laid across the Atlantic, Pacific, and Indian Oceans. This has not gone unnoticed by the Russian military planners. It was reported in the New York Times in October that Russia has stepped up by 50% its submarine patrols near to where these cables are buried. Should this be a worry for companies? Is there a case for companies to reflect a risk of mass cable outages in their strategic decisions on where to locate their storage and computing? Is this another reason for a switch to edge computing?

The cutting of a single submarine cable is common event. The genius of the Internet is to spread the pain so that latency and congestion rise for the days or weeks needed to repair the affected cable that at least allows most functions to manage through. There are also now so many cables that route diversity appears to have coped well with all the events so far, without major disruptions to anyone’s economy or collapse of companies. Can we be so complacent in the future?

There is always a balance to be struck on the probability of an event, its impact, and the cost of doing something about it. The important thing to recognise is that each of these factors is in flux and we certainly need to keep an eye on the trends.

Probability of the event – Some Russian submarines sniffing around submarine cables only signals a capability and not an intent to do damage. The probability of Russian submarines cutting multiple submarine cables under the conditions of the present world order is low. But why the cutting of submarine cables looks a plausible threat in the future is that it is an easy thing to do without getting caught. It offers an option of a “shot across the bows” in a period of rising tension without provoking a nuclear response.

Consequences of an event – This is the trend that generates more concern. Countries and companies are becoming ever more dependent on a reliable Internet. The economic and financial consequences mount from a break in service or sustained acute congestion.  What is likely to be the case is that many companies probably do not know how much of their cloud computing and storage is located across an Ocean. They are therefore in no position to assess the consequences of a multiple attack on underwater submarine cables. The ISP’s probably have a better feel of the massive congestion that would result but have yet to have any experience of many concurrent cuts. It would be like a denial of service attack on a grand scale and sustained for weeks as cable repair ships could only re-splice cables one at a time.

cable repair ship

Cost of doing something about it –  Ultra-cheap long-distance transmission costs, scale economies and cheap energy prices can make a compelling proposition for location of cloud computing and storage on the other side of the world. This suggests that a strategy that avoids the vulnerability of undersea cable breaks will come at a price.

It is doubtful if this reporting of Russian submarines exercising in numbers around our submarine cables will have many companies rushing to reduce their vulnerability to a break of trans-oceanic Internet traffic. Companies must set priorities and the case is not compelling for this to be one of them…at least for now. But those who have responsibility inside their companies for disaster recovery should at least add this threat to the long list they already have for modelling the various disaster scenarios and the effect on their companies. From this more generally modelling of all the threats it must surely make the case for more edge caching of content and more edge computing.

Where the government have it within their gift to encourage the trend towards edge caching and edge computing is to change the conversation on our national broadband objectives (both for fibre and 5G) away from ever higher data speed’s and instead make the conversation about dramatically bringing down “latency”. Getting latency down to a few milli-seconds improves customer experience significantly, makes a reality of the tactile Internet and opens up opportunities for exciting new applications. A market demand for “low latency” is what will make the business case for far more edge caching and edge computing. This will drive a virtuous circle of better scale economies leading to more pervasive deployments. The incidental benefit of all this will be to make the UK less vulnerable to a Russian threat to cut undersea cables upon which today’s Internet is over reliant.

Comments are closed.