The frequency of Fargate Spot interruptions in the Tokyo region has increased, causing resource shortage.
To address this, it is recommended to adjust the grace period for replacing Fargate Spot instances, increasing it from the default 30 seconds to 120 seconds.
Additionally, AWS Fargate now supports AWS Graviton-based Spot compute for ARM architecture.
It is advised to monitor Fargate Spot occurrence status using EventBridge and Cloudwatch Logs.