DataBridge Watch

What Watch does

Watch aggregates operational status from every server in your DataBridge fleet and presents it in a single web-based dashboard. Server health, active ingest lanes, transfer progress, disk utilization, integrity verification results, alerts, and complete audit trails — all in one place, updated in real time.

It connects to the DataBridge backend API via WebSocket. When a server reports a status change, Watch renders it within seconds. When an alert fires, it appears immediately. When a transfer completes and verification passes, the dashboard reflects it without anyone pressing a button.

Watch is not a reporting tool you check at the end of the day. It is the live operational picture your team keeps open on a second monitor or a wall-mounted display, always current, always accurate.

Fleet at a glance

Live server grid, aggregate stats, and recent alerts — the three views you see when you open Watch.

Fleet dashboard

Every server in your fleet, visible at a glance. Online or offline. Active lanes, current throughput, disk utilization, last heartbeat. Color-coded health indicators make it obvious which servers need attention — green for healthy, amber for warning, red for critical. No clicking through individual machines to understand fleet state. One screen, every server, updated continuously.

Transfer tracking

Every storage device is tracked from the moment a label is created in DataBridge Tag through buffer copy, cloud upload, integrity verification, and archival. Watch shows progress percentages, real-time throughput, estimated time to completion, and the current pipeline stage for every active device. When something stalls, you see it immediately — not after an operator reports it an hour later.

Alert management

Alerts fire automatically when something goes wrong. An upload stalls for longer than the configured threshold. A server's disk usage crosses 85%. A network link drops. Integrity verification fails. Each alert carries one of three severity levels — critical, warning, or info — and is pushed to Watch the instant it triggers. Operators can acknowledge alerts, mark them resolved, or escalate to engineering. Every action is logged.

Audit trails

Every status change for every device is recorded permanently. When a label was created, by whom, where. When the device was plugged into the ingest server. When the buffer copy started and finished. When the upload began, its progress checkpoints, when verification passed. Who acknowledged an alert, who resolved it, what notes they left. The audit trail is immutable — entries are appended, never modified or deleted. This is the compliance record for your data chain of custody.

Integrity verification dashboard

SHA-256 manifests are generated during every buffer copy. After cloud upload, ETags are compared against the manifest to confirm every byte arrived intact. Watch surfaces this in a dedicated view: how many devices have been verified, how many are pending verification, how many failed. Drill into any device to see file-level results — individual file paths, expected hash, actual hash, pass or fail. Nothing is assumed. Everything is proven.

Batch tracking

Field teams collect data in batches — a day's shoot, a week's survey, a project phase. Watch groups devices by batch and shows batch-level progress: how many devices have been labeled, ingested, uploaded, verified. Completion percentage for the entire batch. Which devices have been returned to the field team after offload, and which are still at the datacenter. Operations teams use this to coordinate logistics and ensure nothing gets lost between field and datacenter.

Role-based access

Not everyone needs to see everything. Watch supports three roles, each with visibility and permissions tuned to what that team actually needs.

Real-time, not near-time

Watch does not poll. There is no 30-second refresh cycle, no stale data, no “last updated 2 minutes ago” disclaimer. The dashboard connects to the DataBridge backend via a persistent WebSocket, and the backend pushes state changes the moment they occur.

Under the hood, the backend runs on FastAPI with a WebSocket endpoint. MongoDB change streams feed status updates into the socket in real time — when Core reports a lane status change, when an alert fires, when a verification completes, the event propagates through the change stream, into the WebSocket, and onto your screen. Typical end-to-end latency from server event to dashboard render is under one second.

If the WebSocket disconnects — network blip, laptop wakes from sleep — Watch reconnects automatically and fetches any events it missed during the gap. You never have to manually refresh or wonder if what you are looking at is current.

Where Watch fits in the pipeline

DataBridge is five modules working together. Tag labels devices in the field. Core ingests and uploads at the datacenter. Ops gives datacenter operators a local console. Watch is the global monitoring layer — it sees everything, across every server, in real time.

Tag creates the label. Core processes the data. Ops manages the local server. Watch monitors the entire fleet. Direct Line moves the bytes over private infrastructure. Together, they form a complete chain of custody from field capture to verified cloud archive.