The VM pool is why agents start in seconds, not minutes. The platform maintains pre-booted VMs ready for instant assignment. When you spawn an agent, it claims a warm VM immediately instead of waiting for a cold boot.Documentation Index
Fetch the complete documentation index at: https://docs.murmur.dev/llms.txt
Use this file to discover all available pages before exploring further.
How the pool works
The platform continuously maintains a set of warm VMs:- Backfill: When idle VMs drop below
min_idle, new ones are booted - Assignment: When you spawn an agent, a warm VM is claimed instantly
- Return: When an agent finishes, the VM returns to the pool (if under
max_idle) - Scale down: Excess idle VMs beyond
max_idleare destroyed - Hard cap: Total VMs never exceed
max_vms
Pool configuration
Each tenant has apool-config catalog resource (singleton, always named default):
| Field | Type | Description | Example |
|---|---|---|---|
max_vms | int | Hard cap on total VMs across all developers | 20 |
reap_stranded_agents | string | Auto-cleanup mode: enabled, dry_run, or disabled | enabled |
min_idle field controls per-workspace warm VM count.
View current pool configuration:
Monitoring pool health
- Running VMs (assigned to agents)
- Idle VMs (warm, waiting for assignment)
- Pending boots (being provisioned)
- Current configuration
Pool flush
Destroy all warm (idle) VMs. Useful after changing environment config or images:Cost management
Warm VMs consume compute even when idle. Tune these parameters to balance speed vs cost:| Parameter | Trade-off |
|---|---|
Higher min_idle | Faster agent startup, higher idle cost |
Lower min_idle | Slower startup (may need cold boot), lower idle cost |
Lower max_vms | Hard cost cap, but may queue agent starts during peaks |
min_idle of 1–3 is a good starting point for most teams. Agents that finish return their VM to the pool, so active teams naturally keep the pool warm.