BLOG

Simplifying Day-2 Operations with Agent Pools

September 2, 2025
Naveen Chakrapani

Implementing Day-2 Operations such as agent replacement is cumbersome today because every configuration tied to a previous agent must be reconfigured manually. This makes tasks like scaling, retiring agents, or handling failures both error-prone and time-consuming.

To address this pain point, we are introducing the concept of an Agent Pool.

Why Agent Pools?

Instead of binding configurations directly to individual agents, customers can now attach multiple agents to a shared Agent Pool. Configurations such as Environment Templates and Resource Templates reference the pool, rather than a single agent.

This simple shift brings significant operational benefits:

  • Seamless Failover and Replacement
    Add or remove agents from a pool without reconfiguring existing associations.
  • Simplified Day-2 Operations
    Manage scaling, upgrades, and retirements without disruption.
  • Load Balancing
    Distribute load across multiple agents within a pool for higher availability and performance.

Key Capabilities

  • Attach multiple agents to a pool.
  • Associate configurations (environments, environment templates, resource templates) with a pool instead of individual agents.
  • Easily add/remove agents from a pool without breaking existing setups.
  • Continue using existing controls such as agent overrides and project sharing.

Coming Soon: Repository support will be added in a future release.

Conclusion

With Agent Pools, Day-2 operations become simpler, more resilient, and easier to scale. By decoupling configurations from individual agents, customers can:

  • Seamlessly implement failover and agent replacement
  • Add or remove agents dynamically without service disruption
  • Leverage load balancing for better reliability and performance
Tags:
Recent Posts
Part 2: Self-Service Fractional GPU Memory with Rafay GPU PaaS
Self-Service Fractional GPUs with Rafay GPU PaaS
Unlock the Next Step: From Cisco AI PODs to Self-service GPU Clouds with Rafay
Kubernetes v1.34 for Rafay MKS
Dynamic Resource Allocation for GPU Allocation on Rafay's MKS (Kubernetes 1.34)