Cloud-Based Job Scheduler Design
Budget: $750 – $1,500 USD
Looking for a developer with good understanding networking and cloud technologies to develop an agent, central system and job schedular.
Here is what I’m looking for:
• Architecture blueprint showing how the scheduler discovers hardware, tracks resource usage (CPU, RAM, GPU, network), and chooses the best host
• Container-first workflow: Docker images pulled from a registry, started, updated, and retired with zero-downtime rolling updates
• Networking model that keeps service discovery and inter-container communication secure and low-latency
• Automation scripts or IaC (Terraform, Ansible, or similar) that stand up the core components and let me add new nodes quickly
• API or CLI outline so other tools can trigger jobs or query status
• Clear, step-by-step documentation covering setup, scaling, and fail-over scenarios
The scheduler’s primary focus is cloud-grade task automation, and every workload it manages will be a microservice. I’m flexible on platform, so feel free to base the design on Kubernetes, Nomad, Swarm, or a custom Go/Python control plane—just explain why your choice is the right fit for mixed hardware pools.
Here is what I’m looking for:
• Architecture blueprint showing how the scheduler discovers hardware, tracks resource usage (CPU, RAM, GPU, network), and chooses the best host
• Container-first workflow: Docker images pulled from a registry, started, updated, and retired with zero-downtime rolling updates
• Networking model that keeps service discovery and inter-container communication secure and low-latency
• Automation scripts or IaC (Terraform, Ansible, or similar) that stand up the core components and let me add new nodes quickly
• API or CLI outline so other tools can trigger jobs or query status
• Clear, step-by-step documentation covering setup, scaling, and fail-over scenarios
The scheduler’s primary focus is cloud-grade task automation, and every workload it manages will be a microservice. I’m flexible on platform, so feel free to base the design on Kubernetes, Nomad, Swarm, or a custom Go/Python control plane—just explain why your choice is the right fit for mixed hardware pools.