IT
04-12-2025
Cloud Dependence
Cloud Dependence and Systemic Vulnerability
The recent outage of Amazon Web Services (AWS) once again highlights something that has been happening for years: a large part of the global digital economy is concentrated in very few infrastructure platforms.
- Market concentration
AWS, together with Microsoft Azure and Google Cloud, controls most of the world’s cloud computing.
If any one of these providers fails, it implies:
- Interruptions in essential platforms (banks, hospitals, logistics, media, e-commerce).
- Outages in secondary services that depend indirectly on AWS.
- A domino effect in applications that use the cloud only for authentication, databases or storage.
- Fragility of digital chains
Many modern services have no local alternatives or contingency mechanisms.
Even if a company has redundancy, if all its availability zones are on AWS, the dependency remains.
Typical examples:
- Companies that store their backups… with the same provider.
- Systems that depend on external APIs hosted on AWS.
- Applications that only work with centralized authentication or storage.
- Associated risks
- Operational: service loss, business interruption.
- Economic: million-dollar losses for minutes or hours of downtime.
- Reputational: end users lose trust.
- Geopolitical: governments dependent on foreign infrastructure.
- Why it keeps happening
Because the cloud offers enormous advantages:
- Immediate scalability.
- Lower costs.
- Outsourced maintenance.
- Advanced security.
- Fast implementation.
This leads even critical services to move there, increasing dependency.
- Where the discussion is heading
In light of incidents such as the AWS outage, the industry is beginning to rethink how to design infrastructure to avoid single points of failure. The main lines of evolution are:
5.1 Smart multicloud (not just “having several clouds”)
Many organizations say “we use several clouds,” but in reality they use:
- AWS for almost everything, and
- another provider only for backups or marginal operations.
That is not real multicloud.
The trend is to implement portable architectures that allow workloads to be moved between clouds automatically or at least with minimal interruption.
How is this achieved?
- Containers and Kubernetes (decouple the application from the provider).
- Declarative infrastructure (Terraform, Pulumi).
- Databases replicated across clouds.
- Global load balancers that enable failover between providers.
Benefits:
- If AWS fails in a region or zone, the application can be restarted in Azure or GCP.
- It reduces dependency and improves resilience.
5.2 Edge computing: distributing instead of centralizing
Another strong trend is to move part of the processing and storage closer to the user or into decentralized nodes.
Examples:
- Advanced CDNs (Cloudflare Workers, Fastly Compute).
- Local servers in critical businesses (industry, healthcare).
- Regional nodes for IoT or real-time analytics.
Benefits:
- Lower latency.
- Less dependence on the main data center.
- If the cloud goes down, part of the functionality continues operating locally.
5.3 Decoupled and eventual architectures
The idea is to design systems capable of:
- Degrading in a controlled way.
- Continuing to operate partially.
- Reconnecting when the cloud is back.
Examples:
- Apps that work offline and synchronize later.
- Microservices isolated from one another, avoiding cascading failures.
- Use of circuit breakers to prevent external failures from paralyzing everything.
Benefits:
- The organization does not grind to a halt if an external service goes down.
5.4 Evolved hybrid infrastructures
Many companies are returning to a mixed approach:
- Cloud for elasticity and cost efficiency.
- Own infrastructure for critical or high-availability functions.
But unlike the “old” model, it is now integrated with automation and APIs so that both worlds work together.
If you have any questions regarding this topic, please do not hesitate to contact me at +54 9 11 15 2759 1175 or by email at luismatas@jebsen.com.ar
Sincerely,
Luis Matas
IT
IT
December 2025
This newsletter has been prepared by Jebsen & Co. for the information of clients and friends. Although it has been prepared with the greatest care and professional zeal, Jebsen & Co. does not assume responsibility for any inaccuracies that this bulletin may present.