Publication date:
Publication from:
Data sovereignty has stopped being an abstract policy conversation. NIS2, DORA, tighter scrutiny of the US CLOUD Act, and EU customers’ own requirements for subcontractors are turning into concrete line items in technical requirements and tenders. And this is exactly where any business working with European clients — or planning to host infrastructure in the EU — runs into a problem that rarely gets discussed openly: European cloud providers have not yet rebuilt the depth of AWS’s managed services catalog, and the gap is structural, not cosmetic.
A recent industry analysis by DevOps and cloud architect Fedir Kompaniiets (co-founder of Gart Solutions) puts numbers and service-by-service detail behind this gap. Below are the key takeaways — and, more importantly, the practical choice that follows for any company weighing a move off AWS toward European infrastructure, or an entry into the EU market.
AWS in Frankfurt or Dublin is the same full catalog of 200+ managed services available in any other region. The picture among European IaaS-first providers looks very different:
| Provider | Managed services catalog depth |
|---|---|
| AWS (Frankfurt/Dublin) | 200+ |
| OVHcloud | ~45 (the deepest EU catalog) |
| Scaleway | ~30 (the most developer-complete catalog) |
| Hetzner | ~8 (intentionally minimal IaaS) |
One important distinction from the analysis: a “managed service” is not a VM running a Docker container with PostgreSQL inside. It’s a service where the provider itself owns patching, failover, backup, and SLA enforcement. Much of what gets labeled “managed” on smaller European platforms is actually closer to self-hosted with some automation layered on — and the difference in operational load on the customer’s team is substantial.
Managed PostgreSQL/MySQL is the single biggest migration blocker. European clouds are already reasonably solid here: OVHcloud, Scaleway, Exoscale, IONOS, and UpCloud (via Aiven) all offer HA clusters, point-in-time recovery, and connection pooling at a level close to AWS RDS. But none of them has an equivalent to Aurora — the distributed, cloud-native storage tier with true scale-to-zero, instant state rewind (Backtrack), and a Global Database for cross-region replication. Teams that depend on Aurora Serverless for variable workloads currently have no migration path to the EU at all.
This is where the gap is smallest: OVHcloud MKS, Scaleway Kapsule, Exoscale SKS, IONOS, and UpCloud all offer a free control plane and node autoscaling. The weak spot is shared filesystems (RWX volumes): where AWS EFS makes ReadWriteMany trivial, European providers either force customers to overpay for oversized File Storage (a 150 GB minimum at OVHcloud) or require manual NFS setups.
CloudWatch + X-Ray + OpenSearch is a single integrated bill with zero infrastructure operations. No EU provider offers comparable coverage: OVHcloud ships logs (via Graylog) without metrics, Scaleway ships metrics and logs without tracing, Exoscale offers search without metrics. Distributed tracing — an X-Ray equivalent — is essentially absent everywhere. Teams without a dedicated SRE function either keep running CloudWatch for their EU workloads (undermining the sovereignty argument itself) or operate a self-assembled Prometheus/Grafana/Loki stack, which becomes a recurring operational burden.
This analysis points to a very practical fork in the road — for any company, Ukrainian or European, working with regulated clients or simply trying to reduce its legal exposure under the US CLOUD Act.
Option 1. Stay on AWS or another hyperscaler. You get full service depth and zero operational load on your own team — but legal exposure under US law remains, and a path to certifications like SecNumCloud is structurally closed to AWS. For parts of the EU public sector, financial institutions, and defense customers, this is no longer a theoretical risk; it’s a direct contract condition.
Option 2. Move to European IaaS and build the managed layer yourself. This is technically possible: Patroni for HA Postgres, Knative for a Lambda-compatible FaaS layer, a self-run Prometheus/Grafana/Loki stack. But it means your engineering team absorbs the operational work that AWS performs on your behalf — patching, failover, monitoring your own monitoring. For companies without a mature DevOps/SRE practice, this is often a hidden cost that gets underestimated during migration planning.
Option 3. European IaaS plus a managed layer operated by a systems integrator. An SI or MSP takes on the operation of the database, the observability stack, the FaaS runtime, and DR orchestration on top of raw infrastructure from OVHcloud, Hetzner, Exoscale, or others — and sells it as a service with its own SLA. The analysis identifies this as the main growth area for the next few years: European providers aren’t investing in PaaS depth fast enough on their own, so the gap is closing from above — through integrators building the operational layer — rather than from below, through the cloud providers themselves.
For Ukrainian companies, this fork has two additional dimensions.
First, demand from clients themselves. Ukrainian product companies and outsourcing teams working with European customers in fintech, healthtech, or the public sector are increasingly receiving requirements around EU data residency and DORA/NIS2 compliance — regardless of where the development team itself is physically located.
Second, an engineering opportunity. The managed-services gap is demand for exactly the expertise Ukraine’s IT industry has traditionally been strong in: DevOps, SRE, and building reliable self-managed Postgres, Kubernetes, and observability stacks. If the managed layer over European IaaS ends up being built by systems integrators rather than the cloud providers themselves, that’s a concrete market niche where Ukrainian engineering teams already hold a practical advantage.
Before deciding on a migration or an entry into the EU market, it’s worth answering three questions honestly:
The answers will differ for every business — but the gap in managed-services catalogs itself can no longer be dismissed as a technical footnote. It’s already shaping tenders, contracts with regulated clients, and long-term architectural strategy.
This article is based on an industry analysis by Fedir Kompaniiets (Gart Solutions), “The EU Cloud Managed Services Gap.”
