Spanda

Mesh Partition Handling

Partitions are detected when reachable node clusters become disconnected.

Behavior

When a partition occurs:

1. Identify affected entities and clusters
2. Elect a local coordinator (communication role)
3. Apply MeshPartitionPolicy (pause unsafe missions, allow safe missions if configured)
4. Record MeshPartitionReport evidence
5. Sync when reconnected via MeshMergePlan / MeshMergeReport

Types

• MeshPartition, MeshCluster
• MeshPartitionPolicy, MeshPartitionReport
• MeshMergePolicy, MeshMergePlan, MeshMergeReport
• MeshConflict — resolved using decision conflict precedence

CLI

spanda mesh simulate-partition robot-a robot-b
spanda mesh merge-report

API

GET /v1/mesh/partitions · POST /v1/mesh/simulate-partition · GET /v1/mesh/merge-report

Partition mode blocks high-risk mission starts (more restrictive than normal mode).

This site is open source. Improve this page.