Hey folks, been wrestling with a deployment headache this week and wanted to share in case anyone’s hit the same wall. We’re rolling out SonarQube 9.9 LTS for a client’s CI/CD pipeline, hosting it on AWS Fargate. The Docker image (official SonarSource one) pulls and starts, but it consistently fails the ECS health check, cycling tasks endlessly.
Our setup is pretty standard:
- Fargate platform version 1.4.0
- Health check configured to hit `/api/system/status` on port 9000 (HTTP)
- Container gets to “SonarQube is operational” in the logs, but the health check returns 503s after about 2 minutes.
We’ve already ruled out the usual suspects:
* VPC networking and security groups are wide open for the health check source IP
* Container has ample CPU/memory (4 vCPU, 8 GB)
* No obvious errors in the SonarQube logs about DB connectivity (using RDS PostgreSQL)
My hunch is there’s a readiness timing issue with how the internal components bootstrap in this version under Docker. Has anyone else run SonarQube 9.x successfully on Fargate? If so, did you need to tweak the health check path, add a grace period, or modify the sonar.properties inside the container?
For context, we’ve run older versions (8.x) on EC2 without a hitch, so this feels like a platform-image interaction quirk. Any war stories or config snippets would be a lifesaver.
Test early, test often.