Skip to content

Platform depth

Dataplicity is not only a way to open a terminal on a Linux device.

It is an operating layer for the part of a connected product that begins after shipment: identifying the right device and cohort, detecting a meaningful change, gathering evidence, taking a controlled action, and proving the outcome.

These guides show the product in use. They include focused product screenshots, query and selector recipes, production checklists, failure modes, and one end-to-end incident lab.

Dataplicity operational loop connecting identity and scope, signals, evidence, controlled action, and verified outcome
The product depth is in the joins: one fleet identity carried from signal through evidence, action, and verification.

Choose a guide

Logging for a fleet, not a terminal

Centralise device, agent, platform, and external events in one organisation stream. Learn the field-aware query grammar, work through five investigation recipes, turn repeated patterns into throttled webhook signals, and preserve evidence with an incident.

Open the Logging guide →

Run the fleet from one browser workspace

See how inventory, device context, Logging, monitors, Fleet jobs, incidents, Remote Shell, Wormhole, Diagnostics, Agent Log, and the Web CLI divide responsibility while keeping the same fleet identity.

Open the browser workspace guide →

Monitor the cohort, not just the device

Build fleet signals from classes, property tags, custom tags, any/all matching, exclusions, ratios, minimum impacted counts, persistence, and hysteresis. Use worked patterns for ISP cohorts, cold-chain sites, rollout waves, and maintenance.

Open the Fleet monitors guide →

Request evidence across unreliable connectivity

Understand File Transfer snapshots and chunking, design a safe /home/dataplicity/remote handoff directory, build bounded diagnostic bundles into the image, and validate evidence collected over several short connections.

Open the File Transfer guide →

See the features work together

Diagnose the Northstar cold-chain incident is a complete worked lab:

  1. identify the customer, site, device, class, and tag cohort
  2. interpret a cold-chain Fleet monitor
  3. narrow organisation logs around the first signal
  4. choose Remote Shell, Fleet job, or File Transfer based on evidence
  5. inspect per-target outcomes
  6. close only after independent recovery
Dataplicity Fleet jobs showing commands, job state, and mixed per-target progress
The platform keeps partial outcomes visible. One completed job in this view still contains three failed targets.

Capability map

Operational needDataplicity capabilityDeep guide
Search one time window across device and platform activityOrganisation LoggingLogging
Distinguish one failed device from a failed populationFleet monitors + tagsFleet monitors
Keep identity visible while moving between toolsBrowser workspaceBrowser operations
Execute across selected devices and inspect every resultFleet jobsBrowser operations
Collect a prepared artifact from a flapping deviceFile TransferFile requests
Preserve ownership, decisions, and recoveryIncidentsNorthstar lab

Reference remains task-focused

The Platform depth guides explain capability, design, and composition. Use the canonical reference pages for concise task instructions: