IBM Cloud Amsterdam outage exposes serious transparency problem

Cloud providers talk a lot about resilience. Redundancy, failover, business continuity. The language is everywhere in marketing materials and sales conversations. What happened to IBM Cloud‘s AMS3 datacenter on Thursday morning, however, tells a rather different story, and the communication breakdown surrounding it is arguably more troubling than the outage itself.

IBM Cloud’s entire AMS3 facility near Amsterdam went offline for at least four hours after a fire broke out at the NorthC datacenter in Almere, a location sitting just a short distance from Schiphol airport. Fire brigade units from both Amsterdam and Schiphol attended the scene. IBM later confirmed that the facility serves its operations alongside other tenants, and that staff had fully evacuated with no reported injuries. Those are the facts of the physical event, and they are serious enough standing alone.

What compounds the situation considerably is what IBM’s status page showed during those four hours: nothing at all. No incidents. No degraded services. Not a single acknowledgment that an entire European datacenter had gone dark and taken customer workloads with it.

Third-party monitoring services stepped in to fill that silence. StatusGator flagged IBM Cloud as down based on reports from at least ten users, with the final outage report logged at 2325 UTC. Downdetector captured a steady stream of complaints beginning around 0715 UTC and running through until roughly midday. Meanwhile, customers who had already raised Severity 1 tickets, the highest possible priority classification, sat waiting for responses that took hours to come through.

One affected customer described the experience bluntly, noting that useful information only surfaced after reaching out directly to an account manager rather than through any official incident channel. For enterprise cloud support, particularly at the highest severity level, that is simply not an acceptable communication standard.

When IBM’s response finally arrived, it confirmed the fire and described active coordination with emergency services and affected clients. Appropriate, yes, but the hours of silence that preceded it had already done damage to customer trust.

Moreover, this event fits into a wider pattern worth examining. Over the past year, IBM Cloud has reported many serious disruptions, ranging from a severity one outage to individual login failures. More recently, the company also narrowed its Basic Support tier, removing the ability for lower-tier customers to escalate technical support cases through standard portal channels.

Considered together, these developments raise questions that extend well beyond any single fire.