Summary
We should add proper ECS task definition/container-level health checks. This will make container startup failures and runtime health problems easier to detect and debug, and allow ECS and the load balancer to remove unhealthy tasks automatically.
Problem
At present many containers may start without explicit healthchecks. When containers fail to become ready (for reasons such as missing config, slow migrations, failing DB connections, or internal errors) the scheduler and ALB may consider them healthy and route traffic or keep restarting them without clear diagnostics. This increases downtime and makes root-cause analysis harder.
Summary
We should add proper ECS task definition/container-level health checks. This will make container startup failures and runtime health problems easier to detect and debug, and allow ECS and the load balancer to remove unhealthy tasks automatically.
Problem
At present many containers may start without explicit healthchecks. When containers fail to become ready (for reasons such as missing config, slow migrations, failing DB connections, or internal errors) the scheduler and ALB may consider them healthy and route traffic or keep restarting them without clear diagnostics. This increases downtime and makes root-cause analysis harder.