Monitor login

Sign in with a users row (MySQL username + password, or SQLite user + password).

Ops console socket: … spill: —
Active jobs
0
Slots busy
0
Rows (live)
0
Batches
0
Rows/s (avg)
0
Heap peak (proc)
0
Σ wait
0
Workers
0
Alerts
0
leader: — autoscale: — targets: — claims: 0 coord tick: —
Capacity by worker type
—
Control plane
—
Alerts (recent)
—
Autoscale decisions
—
Slots
Capacity trends
Σ waiting
workers
alert rate
rows/s
Throughput (sampled)
rows/s
heap MB
Live data traffic

Speedometer blends row throughput, spill batches, queue wait, and live jobs—one glance at how “loud” the pipeline is. Arc ticks are MB (0–500), evenly spaced along the semicircle: 0, 5, 10, 20, 30, 40, 50, 100, 150, 200, 400, 500.

0%
Idle
Ten situational lenses

Each chart is a short history of one idea about system health—not ten redundant metrics, but ten ways to read the same living organism.

Velocity: are results moving through the pipe?
1 · Flow (rows/s)
—
Backlog: demand waiting before execution.
2 · Pressure (Σ wait)
—
Parallelism: how many fires burn at once.
3 · Concurrency (live)
—
Slots: finite seats—saturation means queue behind.
4 · Runway (slots %)
—
Memory headroom: room to breathe vs heap peak.
5 · Headroom (heap)
—
Fleet: registered workers visible to control plane.
6 · Hands (workers)
—
Alerts: the system’s voice when norms break.
7 · Voice (alerts)
—
Redis: nerve center for queues and bridge.
8 · Nerve (Redis)
—
Spill batches: SQLite absorbing heavy streams.
9 · Sink (batches)
—
Balance: active work vs waiting—supply vs demand.
10 · Equilibrium
—
Timeline
Live processes scroll · max 48 cards · height matches timeline
Process detail
Click a process card.
Queues health
Last health: —
Redis —
Queue wait act ok fail del wrk
Tools