Skip to main content
The Chamber agent is a lightweight Kubernetes controller that connects your GPU clusters to Chamber.

What the Agent Does

The agent runs in your Kubernetes cluster and:

Reports Capacity

Discovers GPU nodes and reports capacity to Chamber

Manages Workloads

Creates and monitors GPU jobs on your cluster

Collects Metrics

Gathers GPU utilization for dashboards

Handles Scheduling

Implements scheduling decisions from Chamber

Architecture

The agent maintains a secure connection to Chamber and executes scheduling decisions locally.

Resilience

Jobs continue running even if Chamber is temporarily unavailable. The agent operates independently once jobs are scheduled. If the connection is lost, running workloads are unaffected and the agent automatically reconnects.

Communication

Outbound Only

The agent only makes outbound connections:
  • No inbound ports required
  • Works behind corporate firewalls
  • Encrypted with TLS

Network Requirements

HostPortPurpose
*.usechamber.io443Chamber communication

Resource Requirements

ResourceRequirement
CPU100m request, 200m limit
Memory128Mi request, 256Mi limit
The agent has a minimal footprint. These defaults work for most clusters.

Security

Permissions

The agent requires these Kubernetes permissions:
rules:
  # Read cluster state
  - apiGroups: [""]
    resources: ["nodes", "pods", "namespaces"]
    verbs: ["get", "list", "watch"]
  # Manage GPU jobs
  - apiGroups: ["batch"]
    resources: ["jobs"]
    verbs: ["create", "delete", "get", "list", "watch"]
  # Read metrics
  - apiGroups: ["metrics.k8s.io"]
    resources: ["pods", "nodes"]
    verbs: ["get", "list"]
The agent does not require cluster-admin. It uses minimal permissions scoped to GPU workload management.

Data Sent to Chamber

DataPurpose
Node inventory (GPU count, type)Capacity tracking
Job metadata (name, status, GPU request)Workload management
GPU utilization metricsDashboards and reporting
The agent does not send container logs, environment variables, or secrets.

Next Steps