London — On Thursday 21 January Seacom was hit by two route failures in Egypt. Almost simultaneously the WACS cable was hit by an outage on the route connecting it to a data centre in London. Russell Southwood looks at what a crisis like this can teach operators and how it provides a good moment to remember the continent's wider unsolved infrastructure problems.
The WACS cable suffered problems slightly earlier than the Seacom outage with a cut in of all places the UK with its route to the London-based data centre where it connects to other routes.
Meanwhile slightly later Seacom was hit by two faults - Northern Trans-Egypt route between Cairo and Alexandria and the other on the Southern Trans-Egypt route on the outskirts of Cairo - cutting off connectivity between Africa and Europe. Apparently both were caused by "civil construction activity" and were not acts of sabotage.
So it was two cables and two breaks in one cable that meant that the usual back-up went down. According to a company statement:"During the two hours and 40 minutes that Seacom experienced a dual failure across Egypt, we were able to route Internet traffic through India. However, many operators in Africa use a basic transmission service that links directly to Europe and they use the west coast cable as a backup," it said.
"With the west coast experiencing an outage at the same time, international connectivity at many of these service providers failed or was degraded while we worked to repair the faults in Egypt."
It is always easiest to be wise after the event and it is understabdable that no-one can have believed that there would be two near simultaneous outages on the West and East coast. But this is the Gods way of telling you to pay attention.
Only having "fallover" back-up on one cable leaves you vulnerable: the more "fallover" arrangements, the better. Since there are now three major cables on the West coast and five on the east coast, this cannot be impossible to arrange. It also highlights why the ACE extension to South Africa continues to be important.
This week's failure on the international fibre routes to the continent provides a sharp reminder that while infrastructure has improved immeasurably over the last decade, much still needs to be done.
The "much to be done" falls into three categories:
1. Countries that have pretty extensive fibre networks both in relation to their geography and where people live... BUT and it's a big but... The quality of network reliability remains poor, particularly in the more remote or busier parts of the network. Africa's data needs continue to increase and the next spike in demand will come with 4G. Too many networks are not "fit for purpose" when it comes to dealing with high volume data demand.
2. Remote and not so remote areas that have little or no coverage: These remain a combination of areas that are market addressable and are also going to require subsidy. With a handful of notable exceptions (eg Ghana, Lesotho, Uganda), African Governments, Universal Service Agencies and regulators are asleep at the wheel on this one. Money is not really the issue for the regulators have collected large amounts of money from operators. In most cases, this money has not been spent.
3. Countries where there is little or no fibre network coverage: These are a combination of two different types of countries: The first category are those that are still beset by civil war or other disturbances including countries like Central African Republic, Somalia, Guinea Bissau and South Sudan. The second category are those where the countries are currently so poor that it makes building a fibre network hard to justify without external, non-commercial funding. These include countries like DRC, Guinea, Liberia and Sierra Leone.
In other countries, there are large geographic areas - usually in the north of a country - where there is no service or only limited networks. These include regions like the northern parts of Ghana, Nigeria and Uganda. Where there is network access, it is expensive and unreliable and no market, however small, thrives in these circumstances.
It has been gratifying to see a company like Viettel build a nationwide fibre network in Mozambique and commit to do the same in Tanzania. But in the main, the unfinished infrastructure tasks in Africa are insufficiently attractive in market terms to have operators want to do this all by themselves.
A country like Sierra Leone has considerable mineral wealth and with proper management of resources could be considerably wealthier in ten years time. But it needs a national fibre network both to deal speedily with the consequences of ebola and to build a wealthier and healthier future.