HarperFast / HarperFast/harper-pro
Add replicated:true support to restart operation
- Dominant language
- JavaScript
- Stars
- 3
- Forks
- 0
- Avg merge
- 1d 21h
- Merged PRs (30d)
- 80
Description
## Request
The `restart` operation (full node restart) does not support `replicated: true`, making it necessary to send a separate `restart` request to each node in a cluster individually.
The `restart_service` operation has `replicated` support (via `replication.sendOperationToNode`), but `restart` does not — see `core/bin/restart.ts`.
## Proposed
Add `replicated` field handling to the `restart` operation, analogous to how `restart_service` handles it: iterate connected nodes and dispatch the operation to each before/during local restart.
```json
{
"operation": "restart",
"replicated": true
}
```
**Note:** Full process restart is more disruptive than `restart_service` — consider whether to restart nodes sequentially (rolling) vs. in parallel.
🤖 Filed by Claude on behalf of Kris.
Contributor guide
Assessment
This issue has not been assessed yet.