Glossary
Observability and IT asset management glossary
Short, plain definitions of the terms used across monitoring, reliability and IT asset management. Where VerOps implements the idea, the link goes to the documentation that explains how.
Observability
- Observability
- The ability to explain what a system is doing from the data it emits, usually traces, metrics and logs, without adding new code for each question. Read more
- APM
- Application performance monitoring: measuring the latency, errors and throughput of application services and the calls between them. Read more
- Distributed trace
- The record of one request as it passes through several services, made up of spans that share a trace ID. Read more
- Span
- One timed operation within a trace, such as an HTTP call or a database query, with a start time, a duration and attributes.
- Service map
- A diagram of services and the calls between them, built automatically from trace data. Read more
- Flame graph
- A view of a trace or profile in which each bar is an operation and its width shows how much of the total time it took.
- OpenTelemetry
- An open-source, vendor-neutral standard and set of SDKs for producing traces, metrics and logs. Read more
- OTLP
- The OpenTelemetry Protocol, used to send telemetry from SDKs and Collectors to a backend over HTTP or gRPC. Read more
- OpenTelemetry Collector
- A process that receives, processes and forwards telemetry, often run on each host or in each cluster. Read more
- Resource attributes
- Key-value pairs that describe where telemetry came from, such as service.name, host.name or k8s.pod.name. Read more
- Cardinality
- The number of distinct values a label or attribute takes. High cardinality, such as a user ID on a metric, multiplies storage and slows queries. Read more
- Rollup
- A pre-aggregated copy of fine-grained data at a coarser interval, so long time ranges can be charted quickly.
- Real user monitoring
- Measuring performance and errors in the browsers of real visitors, as opposed to scripted tests. Read more
- Core Web Vitals
- Google’s user-experience metrics for web pages: Largest Contentful Paint (loading), Interaction to Next Paint (responsiveness) and Cumulative Layout Shift (visual stability). Read more
- Synthetic monitoring
- Running scripted browser journeys or API requests on a schedule to check that an application works, whether or not anyone is using it. Read more
- Private agent
- A test agent you run inside your own network, so internal applications can be tested without exposing them. Read more
Network
- SNMP
- Simple Network Management Protocol, used to read status and counters from switches, routers, firewalls and other devices. Version 3 adds authentication and encryption. Read more
- SNMP trap
- A message a device sends on its own, usually to UDP port 162, when something happens, such as a link going down. Read more
- Traceroute
- A test that lists the network hops between a source and a destination, with the latency to each.
- MTU
- Maximum transmission unit: the largest packet a network path carries without fragmenting it. A mismatch can cause hard-to-diagnose failures.
- Agent mesh
- A set of network agents that test each other, giving latency and loss between every pair of locations. Read more
Reliability and response
- SLI
- Service level indicator: a measure of user-facing behaviour, such as the share of requests that succeed or finish within 300 ms.
- SLO
- Service level objective: a target for an SLI over a period, for example 99.9% of requests succeeding over 30 days. Read more
- Error budget
- The amount of failure an SLO allows. At 99.9% over 30 days, the budget is about 43 minutes of full outage. Read more
- Burn rate
- How fast the error budget is being spent, relative to the rate that would spend it exactly by the end of the period. Read more
- Alert silence
- A rule that mutes matching alerts for a set time, such as a maintenance window. Read more
- On-call schedule
- A rotation that decides who is responsible for responding to alerts at a given time, with overrides for swaps and holidays. Read more
- Escalation policy
- The order in which people or teams are notified when an alert is not acknowledged in time. Read more
- MTTA and MTTR
- Mean time to acknowledge and mean time to resolve: the average time from an alert to someone taking it, and to it being fixed.
- War room
- A shared space for one incident that holds its timeline, evidence and participants. Read more
- Runbook
- A documented procedure for a known task or failure. In VerOps, runbooks can be built as executable workbooks. Read more
- Release gate
- An automated check that a deployment pipeline waits on, continuing only if the check passes. Read more
IT asset management
- IT asset management (ITAM)
- Tracking the hardware and software an organisation owns through its life: purchase, assignment, use, maintenance and disposal. Read more
- Asset inventory
- The list of devices and the software on them, ideally collected by agents rather than kept by hand. Read more
- Software normalisation
- Grouping the many executables and package names of an application into one product, so it can be counted and licensed as one. Read more
- Entitlement
- What a licence agreement allows you to use: a metric (per device, user, core or site), a quantity and a term. Read more
- Licence position
- The comparison of entitlements with actual consumption, showing whether you are under-utilised, compliant or over-deployed. Read more
- True-up
- The periodic reconciliation with a vendor, in which you pay for extra use or adjust a subscription to what is really needed. Read more
- Software utilisation
- How much an installed application is actually used, measured over time rather than by whether it is installed. Read more
- Shadow IT
- Software or services in use without the knowledge or approval of IT. A discovery survey is the usual way to find it. Read more
- Patch posture
- How up to date a device is with operating-system and software updates. Read more
- Straight-line depreciation
- Reducing an asset’s book value by the same amount each period, from purchase cost down to its salvage value over its useful life. Read more
- Enrolment token
- A short-lived, limited-use secret that lets a new agent register once and receive its own device credential. Read more
Platform and AI
- Tenant isolation
- Keeping each customer organisation’s data separate, so no query or user can read another organisation’s data. Read more
- RBAC
- Role-based access control: granting permissions to roles, and roles to people, rather than to each person directly. Read more
- LLM-as-judge
- Using a language model to score another model’s output against stated criteria. Useful for spotting drift, but a judgement rather than proof. Read more
- Model Context Protocol (MCP)
- An open protocol for connecting AI applications to tools and data sources. Read more
Read more in the docs
Try it on your own systems
Create a workspace, or tell us about your environment and we will walk you through it.