Browse documentation
Workspace upgrade policy
Decide whether existing runs move to a new deployment before their next turn.
View MarkdownEach deployment declares how it affects existing durable runs.
Automatic upgrades
runtimeConfig:
workspaceUpgradePolicy: automaticBefore the next turn on an eligible existing run, Salambo:
- Claims the run upgrade lease.
- Checkpoints the current workspace.
- Creates a sandbox from the active deployment.
- Restores the existing workspace.
- Validates the target runtime.
- Atomically switches the run to the new sandbox and deployment.
- Executes the requested turn.
- Saves a post-turn checkpoint before reporting completion.
The old sandbox remains authoritative until the switch succeeds.
New-run-only upgrades
runtimeConfig:
workspaceUpgradePolicy: new-run-onlyExisting runs remain pinned. Only new runs use the deployment.
Workspace preservation rule
The new deployment's seed is never merged into an existing run. Existing workspace files are restored byte-for-byte from the run checkpoint.
Failure behavior
| Failure | Result |
|---|---|
| Checkpoint fails | Old sandbox remains authoritative |
| New sandbox creation fails | Old sandbox remains authoritative |
| Restore or validation fails | Replacement is discarded and retry remains possible |
| Atomic switch succeeds | New sandbox becomes authoritative and old compute is cleaned up |
| Post-turn checkpoint fails | Turn fails instead of claiming durable completion |
Use new-run-only for changes that should not affect an existing session. Use automatic when compatible runs should adopt the latest deployment while preserving their workspace.