Some compute nodes showed performance issues. These got drained sequentially to avoid impact on workloads. Full performance has been restored.

Currently investigating.

We are currently investigating reported performance issues with compute in our Region HAM1.