Search the practice set
275 questions, 991 terms and 600 topics in 30 areas.
60 results for “Device Telemetry at Scale”
Edge Aggregation
Summarising or filtering telemetry at the device or gateway so that only what is needed leaves, which is usually the difference between a viable and an unaffordable design.
Device Group Rollout
Applying configuration or firmware to a fleet in ordered cohorts with health checks between them, because a bad change applied to everything at once has no remedy.
Google Maps and Planetary-Scale Spatial Serving
Map serving is fast because almost nothing is computed on request — the world is precomputed into a pyramid of tiles, and space is indexed onto a one-dimensional curve.
Inference Telemetry
Recording the full context of each model interaction — inputs, outputs, tokens, latency, model version and evaluation scores — so quality and cost can be investigated.
Paved Road Telemetry
Instrumenting the platform's own usage — where teams succeed, where they stall, where they leave — so its roadmap is driven by evidence.
Scale Cube
A model describing three independent axes of scaling — cloning, functional decomposition, and data partitioning — each addressing a different limit.
Telemetry Cost Management
Controlling observability spend through sampling, retention tiering and cardinality limits without losing diagnostic capability.
Telemetry Sampling
Keeping a subset of traces or events to bound observability cost, chosen so the ones that matter survive.
Telemetry Spend Ratio
Observability cost as a proportion of the infrastructure it observes, used as a tripwire for a category that grows silently.
WhatsApp's Small-Team Scale
WhatsApp served hundreds of millions of users with a few dozen engineers by matching one technology choice precisely to the workload and refusing to add anything else.
Zoom's Pandemic Scale-Up
Zoom grew from around 10 million to over 300 million daily meeting participants in roughly three months, absorbed by a hybrid architecture and a distributed media routing design.
A/B Partition Update
Writing firmware to an inactive partition and switching on next boot, so a failed update falls back to the previous image rather than bricking the device.
Ambassador Pattern
A helper process that handles outbound network communication on behalf of an application, adding retries, routing, security and telemetry without changing it.
Approximate Nearest Neighbour Index
An index that trades exactness for speed when finding similar vectors, making large-scale semantic search feasible.
Business Metric Instrumentation
Emitting metrics for business outcomes — orders, payments, signups — alongside technical telemetry, so incidents can be detected and prioritised by impact.
Cell-Based Architecture
Partitioning a service into complete, independent copies of itself, each serving a subset of customers, so a failure is bounded to one cell.
Client Runtime Constraint
The properties of the browser or device you do not control — CPU, network, version, extensions — which make the client a distributed system component rather than a rendering surface.
Clock Drift
Device clocks diverging from real time, which corrupts ordering, expires certificates early or late, and produces data that appears to arrive from the past.
Compute Tier Placement
Deciding which of on-device, on-premises, metro, regional or central compute a workload belongs to, from its latency, data volume and autonomy requirements.
Contract Verification Gate
A provider's pipeline stage that replays every consumer's recorded expectations and fails the build if any would break.
Data Plane Proxy
The per-workload proxy that actually carries mesh traffic, applying mTLS, retries, timeouts, routing and telemetry outside the application.
Hardware Backed Credential
A private key generated inside a secure element and unable to leave it, so device identity cannot be copied off the device.
Hardware Security Module
A tamper-resistant device that generates and stores keys and performs cryptographic operations without the key material ever being extractable.
Horizontal vs Vertical Scaling
Adding more machines versus making one machine bigger — and the fact that vertical is underrated for stateful tiers.
In-Country Processing
Keeping data within a jurisdiction across every path it takes — including backups, logs, telemetry, support access and the disaster recovery region.
Incident Severity Levels
A small, agreed scale of incident severity that determines response, escalation and communication without requiring debate during the event.
NAT Gateway
A managed device that lets instances in a private subnet make outbound connections without being reachable inbound.
Observability
The property of being able to answer new questions about a system's internal state from its external outputs, without shipping new code.
Port Exhaustion
Running out of available source ports for outbound connections through a NAT device, causing new connections to fail while everything appears healthy.
Reported Versus Desired State
Holding two representations of a device — what it last said it was, and what it should be — with the gap between them driving reconciliation.
Role Explosion
The proliferation of narrowly-scoped roles that occurs when RBAC is used to express rules that actually depend on context.
Service Mesh
An infrastructure layer of sidecar proxies that handles service-to-service networking — mTLS, retries, timeouts, routing, telemetry — outside the application.
Severity Levels
A predefined scale of incident impact that determines who is woken, how fast, and what process applies.
Shopify's Pods and Modular Monolith
Shopify handles Black Friday scale with isolated pods — complete stacks each serving a subset of merchants — while keeping the application itself a deliberately modular monolith.
Wide-Column Store
A store organised as partitions of sorted rows, designed for very high write throughput and predictable single-partition reads at large scale.
Zero Touch Provisioning
A device obtaining its identity and configuration on first boot without a human configuring it, which is the only approach that scales past a few hundred units.
A service autoscales on CPU. During incidents it never scales out, even as latency triples. Why, and what would you scale on instead?
Why CPU never triggers The service is latency bound, not compute bound. Its threads are waiting — on a database, a downstream API, a lock, or a connection pool
An OTA firmware update has bricked 3% of a 50,000-device fleet. What do you do, and what should have prevented it?
Halt immediately, then determine the correlation Stop the rollout before anything else. Then find what the 1,500 have in common: hardware revision, previous fir
After a five-minute network outage at your ingest tier, 200,000 devices reconnect simultaneously and the authentication service collapses. How do you fix this?
The mechanism Every device detected the disconnection at roughly the same moment and retried immediately. Reconnection is the most expensive operation in the fl
Three teams keep breaking each other's services with API changes. Someone proposes contract testing. How do you introduce it without a six-month programme?
Start with one pair, not with the estate Pick the pair of services that has broken each other most recently. Instrument that one relationship end to end: the co
You inherit an IoT fleet of 80,000 devices whose client certificates all expire in fourteen months. What do you do?
Establish whether rotation is even possible The first question is whether the current firmware can obtain and install a new certificate remotely. If it cannot,
Device Telemetry at Scale
Deciding what a device sends, how often, and what is aggregated before it leaves.
Contract Testing at Scale
Keeping dozens of services compatible without an environment that runs all of them.
Device Identity
Identifying a thing rather than a person, and rotating a credential you cannot type.
Device Provisioning
Getting identity and configuration onto hardware at manufacture or first boot.
Platform Telemetry
Instrumenting the platform itself: usage, friction, and where teams leave the paved road.
Telemetry Cost
Observability bills that rival compute, and where to cut without going blind.
Client Feature Flags
Flag evaluation on a device you do not control, and the flicker and staleness it brings.
Data Governance & Semantics
General material on ownership, meaning, quality and control of data at enterprise scale.
Edge Compute Topologies
Regional, metro, on-premises and on-device, and what each tier is genuinely for.
Edge Security & Attestation
Secure boot, hardware roots of trust, and proving what is running on a device.
Enterprise Architecture
General material on architecture at portfolio and estate scale.
Horizontal vs Vertical Scaling
Scale out for stateless, scale up first for stateful.
OTA Updates
Updating firmware over a flaky link, with rollback, and without bricking the device.
Observability Cost
Telemetry bills, cardinality control and retention tiering.
Peak Event Readiness
Freeze, pre-scale, shed order, warm caches and rehearse.
SOLID
Five design principles, two of which scale beyond the class.
Serverless
Scale to zero, per-request billing, cold starts and connection limits.
Service Mesh
Sidecars applied estate-wide, and the scale at which that pays.