The first fifteen days were the impact; the next fifteen were something else. They were the test of whether all of that had been an accident or the beginning of a pattern, and the answer came quickly.
The systems began to recover, but never completely. In AWS internal dashboards, some regions turned green again, others yellow. Some never stopped being red. It wasn’t a total blackout, but a kind of permanent instability, as if the infrastructure were trying to breathe under pressure. And then the drones returned. Not with the spectacle of the first attack, but with something much more effective: persistence. The Bahrain region suffered interruptions again. Not always from direct impacts, but from nearby activity, explosions in the surroundings, interference that forced systems to shut down as a precaution. It was the second time in the same month, and that changed the interpretation of everything. It was no longer an isolated incident; it was a clear sign that digital infrastructure had entered the conflict’s playing field. The companies themselves began to assume it, AWS explicitly recommended that its customers migrate workloads to other regions, something that normally only happens in extreme scenarios. Within technical teams, the language also changed; people no longer spoke of “incidents,” but of an “unstable zone.”
Meanwhile, the war outside the datacenters escalated to another level. Attacks began to systematically target energy infrastructure. Refineries, gas plants, logistics terminals in Qatar, Saudi Arabia, Kuwait, and the Emirates. It wasn’t coincidence; it was military logic.
Energy doesn’t just power vehicles or factories; it is the invisible foundation of everything else. Without energy there is no network, without network there is no data, and without data many economies simply stop functioning. At the same time, the Strait of Hormuz, one of the most critical points in global trade, began to come under strain. The flow of hydrocarbons decreased, prices rose, and with them, the cost of maintaining any energy-intensive infrastructure, including datacenters.
In those same days, another element came into play: the conflict ceased to be contained. The Houthis in Yemen began launching missiles toward Israel, expanding the geographical scope of the war. Iran attacked U.S. bases in Saudi Arabia. The United States deployed reinforcements, including an amphibious group with thousands of marines in the region. The map was no longer local, it was becoming regionalized, and that had a direct consequence for digital infrastructure. Each new actor implied new attack vectors, new risk zones, new routes that could be affected.
Under the sea, the problem began to take shape. There were no major confirmed cuts to submarine cables in these days, but something more subtle. Strategic projects began to be delayed, logistical operations were halted. The 2Africa cable, one of the most important in the world, began to accumulate uncertainty. It wasn’t necessary to cut a cable to generate impact; it was enough to prevent it from being maintained. At the same time, another layer of conflict appeared, invisible but just as real: cyber warfare.
Systems down, internet outages, digital offensive operations—at times, parts of the Iranian network were practically in the dark, and meanwhile, in the rest of the world, alerts increased because if anything had become clear, it was that the next logical step was not only to attack the physical, but to combine it with the digital.
Inside the datacenters, the feeling was different from that of the first days—it was no longer surprise, it was wear. Teams working continuously, systems running in degraded configurations, decisions that previously required weeks now made in hours: moving workloads to Europe, redistributing traffic, accepting higher latencies, prioritizing availability over efficiency. The concept of “high availability” was beginning to be redefined in real time.
Outside, the economic impact began to accumulate. The production of critical materials such as gallium stopped due to energy problems. Key industrial facilities halted, technology companies began to declare force majeure situations. It wasn’t just a regional problem; it was a chain, and each link affected the next. The most unsettling thing was not what had already happened, but what was starting to seem evident. No one was trying to completely destroy the infrastructure, only to keep it under constant pressure—enough for it to never be reliable, enough to force the rest of the world to adapt.
As that pressure increased, in military offices another possibility began to be discussed—one that would completely change the scenario: a landing. Not as a distant hypothesis, but as a real option on the table. If that happens, the objectives will cease to be selective, and energy, logistics, and digital infrastructure will become a direct part of the battlefield.
In the next chapter we will see exactly that—what would happen if the war crosses that threshold, and why, in that scenario, datacenters would cease to be collateral damage and become priority targets.
Fuentes
Tom’s Hardware (2026) — interrupciones continuas en la región AWS Bahrain
DataCenterDynamics (2026) — impacto sostenido en datacenters por el conflicto
Mordor Intelligence — ecosistema de datacenters y conectividad en Oriente Medio
PwC — crecimiento del mercado de datacenters en Oriente Medio