The unacceptable fragility of infrastructures

This is a text automatically translated from Italian. If you appreciate our work and if you like reading it in your language, consider a donation to allow us to continue doing it and improving it.

The articles of Cassandra Crossing I'm under license CC BY-SA 4.0 | Cassandra Crossing is a column created by Marco Calamari with the "nom de plume" of Cassandra, born in 2005.

Third and final chapter on the “End of the World.

This article was written on March 12, 2024 from Cassandra

Cassandra Crossing 578/ The unacceptable fragility of infrastructure

The conclusion of the “End of the World” trilogy. How likely is a total collapse of the Internet? And maybe not just the Internet. And why the hell hasn't the situation improved since the 1990s?

The 90s were still new when a then inexperienced Cassandra read a very interesting article about the fragility of the technological infrastructure of the Internet.

[Author's note This time not even the Internet Archive was able to come to Cassandra's rescue; she looked for the article that appeared on a site from the early 1990s, but was unable to find any trace of it. Readers will therefore have to trust and rely on the memory of their favorite prophetess, or better yet help her, by finding him. Ah, and since this is the last installment of a trilogy, if you haven't already done so, read it Before and the second episode.]

The article didn't actually distinguish between Fragility and Unreliability, confusing the two terms, which are actually different. But let's go in order.

It was said of "Most important node on the Internet”, today we would call it a MIX; that is, a point where the various geographic connectivity providers (not connections last mile) had found it convenient to install their own router, connected to all the others, in order to route packets from one provider to another with maximum efficiency.

These types of measures have been the ABC of the Internet for decades now, but at the time they were solutions artisanal to new problems, problems caused by an unresolved separation between the different areas of NSFnet, and by the heated debate on the routing of commercial traffic on academic networks.

Back then the Internet had more or less 200,000 users, and only universities and large companies had it. For private individuals it was just a dream.

Be that as it may, these vital infrastructures of the Internet were built wherever they could, often without being able to adequately invest in the site.

So it was that someone noticed that the node in question was located in one of the garages of a large multi-storey car park, and wondered what would have happened if the driver of a large sedan had made a mistake in shifting into reverse, and had thus broken through the obviously thin side wall of the box, squashing the routers installed there.

Easy prediction, the traffic of a good part of the Internet at the time would have been blocked, and in the absence of the automatic routing between autonomous systems, still to be invented at the time (forget this last consideration if it seems Arabic to you) a good part of the computers connected to the Internet would have ended up offline.

The thing reached a technical journalist who made a good deal out of it piece, halfway between a technical article and a note of colour. The transience of the web then made it disappear.

Fast forward thirty years and we come to today.

The Internet has grown by 4 orders of magnitude to 2,000,000,000 users, and all noteworthy routers are well kept in specialized data centers, automatic BGP routing exists and is active, the Internet has conquered the radio spectrum and space , and a thousand thousand things and people monitor and protect the traffic of the Internet.

But the Internet has become a global and vital resource, and this has opened it up to very different risks; OGAFAM's commercial choices, terrorist attacks, wars, even being a theater of war itself.

In fact, the modern network shares incredible fragilities with other communication infrastructures, with the electricity distribution network and with the fuel distribution network (oil pipelines, gas tankers, terminals).

All it takes is a single act of cyber warfare, such as launching a fake update on satellite modems, such as a software attack on BGP routing systems, or even an accident such as a ship's anchor drifting and severing a bundle of submarine cables, or perhaps an attack on a terminal of the same cables, to cause big trouble to the Internet. All things already seen, for example, in the case of oil pipelines.

It is very easy to think of the potential of a well-planned and coordinated attack on all the submarine cables of a continent, or of the entire planet.

It is equally easy to think of a "software" attack on the delicate and partially manual mechanisms that keep the traffic of the entire network effective and balanced.

Two simple recipes to cause a collapse of the entire Internet.

And if an Internet collapse occurred at the same time as other acts of asymmetric warfare, wouldn't we risk finding ourselves in a situation just as dangerous, from the point of view of civilization, as a global thermonuclear war, which could last a long time or become permanent?

These are not exaggerations; as the inconveniences and small actions of asymmetric warfare of recent years have well demonstrated, the technological threads of the fabric of our civilization are thin and delicate; it doesn't take much to tear them away, the will of a single nation-state, an economic potentate, a criminal organization may be enough.

And no one has studied whether a multi-domain damage to the technological infrastructure complex could collapse, for a period long enough to be devastating, multiple interconnection systems simultaneously; Network, transport, energy distribution, logistics services, all together.

A collapse of the planet's technological system cannot be ruled out, because the couplings and cascade effects between different domains of technology are little studied and even less understood.

On the other hand, modifying "fragile" technical choices of the past, creating new resilient infrastructures, would require decades that we do not have, as well as capital and resources that simply do not exist in the world.

Black Swans? Double black swans? Flocks of black swans, lovingly raised?

Food does not come out of the earth on its own, it does not arrive on our tables on its own feet. Energy is not created by activating a contract or filling up. Water doesn't come out of the taps on its own. And let alone bits come and go on their own between computers, connected IoT and industrial objects.

How likely is a systemic collapse artificially induced of the planet, how long would it last, what consequences would it have, how many people would die?

And what could we do to avoid it?

Already …

Marco Calamari

This tag @loyal alternatives is used to automatically send this post to Feddit and allow anyone on the fediverse to comment on it.

Join communities