AWS

Amazon EKS now supports reverting the control plane to the previous Kubernetes version within seven days of an upgrade. This safety net allows teams to quickly recover from issues caused by problematic updates during in-place cluster upgrades. The feature directly reduces the operational risk associated with version transitions.

• Control plane rollbacks are supported within a 7-day window post-upgrade.
• In-place upgrade risks are mitigated by this quick recovery mechanism.
• Teams can revert to the prior version if the new one causes issues.
• No need to rebuild clusters from scratch for recent failed upgrades.

Airbus has selected Scaleway as its sovereign cloud provider, prioritizing immunity from non-European legislation alongside technical performance. This move complements its existing multi-cloud strategy rather than replacing hyperscalers like AWS. The decision highlights a broader industry trend where sovereignty claims are becoming a decisive factor in vendor selection, extending beyond major providers to smaller US SaaS vendors.

• Airbus treats legal sovereignty as a scored criterion equal to technical capability in cloud tenders.
• Selection of Scaleway complements rather than replaces existing AWS multi-cloud architecture.
• Sovereignty requirements are expanding to include smaller US SaaS vendors, not just hyperscalers.
• Practitioners must verify that claimed sovereignty controls are enforceable and auditable.

A maintenance error involving fiber optics caused an outage in Azure California for nearly five hours. The incident immediately impacted 27 distinct services, highlighting the fragility of physical layer dependencies in cloud infrastructure. This event underscores the risks associated with routine operational changes in critical connectivity paths.

• Fiber maintenance errors can cause immediate, widespread service disruptions.
• 27 Azure services were affected during the near-five-hour California outage.
• Physical layer changes require rigorous validation to prevent cascading failures.
• Outages often stem from operational mistakes rather than complex software bugs.

AWS has enabled multi-node scaling for Amazon Timestream for InfluxDB 3 Enterprise, allowing clusters to expand to 15 nodes with distinct roles for ingestion, querying, and compaction. The update provides guidance on choosing between vertical and horizontal scaling strategies to balance stability and cost. This capability lets teams isolate workload components to better match specific demand patterns.

• Clusters now support up to 15 nodes with dedicated roles for ingestion, query, and compaction.
• Separating roles allows right-sizing individual components rather than scaling the entire cluster.
• Guidance is provided for deciding when to scale vertically versus horizontally.
• Isolating workloads helps optimize both system stability and infrastructure costs.