Manager
Health metrics for the manager service, which handles the user-facing API, GUI, and admin functions on port 3000. Shows service status, uptime, memory usage, and runtime version information. There is only one manager instance in the architecture.
Fields & columns
| Name | Description |
|---|---|
Uptime | How long the manager process has been running since last restart |
Started | Timestamp when the manager process started |
App Version | The deployed mipo application version |
Node.js | Node.js runtime version powering the manager |
Heap Used | V8 JavaScript heap memory currently in use (managed by garbage collector) |
Heap Total | Total V8 heap memory allocated from the OS |
RSS | Resident Set Size — total OS-allocated memory including heap, stack, and native buffers |
External | Memory used by native buffers and extensions outside the V8 heap (e.g., file I/O buffers, crypto operations). Usually small relative to heap. |
Platform | Operating system the manager is running on (e.g., linux) |
Architecture | CPU architecture (e.g., x64, arm64) |
Gotchas
- RSS includes shared libraries so it will always be higher than Heap Total — this is normal.
- Unlike ingest, there is only one manager instance. If it goes down, the GUI and admin API are unavailable.
- Manager bridges both the external network (serves GUI and API) and internal network (accesses the database cluster via the config/results/jobs façade pools).
API calls (1)
| Method | Path | Description |
|---|---|---|
| GET | /api/health/services/manager | Fetch manager health status and metrics |
Related
- Overview — Architecture diagram showing the manager in context
- Ingest — The other Node.js service — handles scanner API traffic
- Database Status — Manager connects to the Config database