Glossary

AI agent observability is the practice of collecting, correlating, and analyzing telemetry data emitted by autonomous AI agents and multi-agent systems.

AI log analysis uses machine learning and large language models to parse, correlate, and explain log data so engineers can find root cause and meaning faster.

AI log monitoring applies machine learning to log streams to detect anomalies in real time and reduce the alert noise of static, threshold-based monitoring.

AIOps is short for Artificial Intelligence for IT Operations. It uses machine learning, AI techniques, and big data to improve operations. Examples of AIOps include correlating events and anomaly detection.

Multiple servers spread across different cities or countries share a single IP address. When a user sends a request to that specific address, the network automatically routes their traffic to the closest available server.

Learn about Apache Iceberg, the open-source table format for massive analytic datasets. Discover its ACID transactions, schema evolution, and performance.

Autoscaling is a feature of cloud computing that automatically adjusts the amount of compute resources needed based on demand.

Amazon Kinesis is a set of cloud services for real-time streaming data collection, processing, and analysis. It enables ingesting and processing large-scale data streams from many sources on a continuous basis.

Batch ingest collects large amounts of raw data from source systems and moves it into a central storage repository all at once on a set schedule.

Bitpacking is a column compression technique that stores integers using only the bits needed for the value range in a block, shrinking storage and accelerating analytic queries on columnar data.

Comprehensive guide to bot management implementation covering detection strategies, business value, technical challenges, and best practices for handling automated traffic across web properties, APIs, and mobile applications at scale.

Cardinality matters because it directly affects how you model, query, and store data.

Cardinality, in the context of metrics systems and observability pipelines, refers to the number of unique combinations of label or tag values attached to a metric or event stream.

Comprehensive guide to CDN log analytics covering performance optimization, security monitoring, and multi-provider challenges. Learn implementation strategies for Cloudflare, AWS CloudFront, Akamai, and Fastly with schema normalization, streaming pipelines, and cost optimization best practices.

CDN analytics involves the systematic collection and analysis of content delivery network logs and metrics to monitor performance, troubleshoot delivery issues, optimize costs, and make informed traffic routing decisions across global audiences and multiple CDN providers.

CDN capacity planning forecasts traffic demand and provisions edge and origin resources so a content delivery network absorbs peaks without congestion or wasted spend.

Comprehensive guide to CDN event delivery covering multi-provider integration challenges, schema normalization, delivery modes, and best practices for streaming log data from CDN edge networks to analytics infrastructure for security monitoring and performance analysis.

CDN incident response is the process teams use to detect, diagnose, and resolve content delivery network problems like edge outages, latency spikes, and cache failures.

Content delivery network monitoring is the process of collecting and analyzing data from the servers that deliver your website or application to users. When someone requests a video or a webpage, that request usually hits a CDN server near them instead of your central data center.

CDN performance measures how effectively a content delivery network serves digital assets to end users.

CDN switching is the practice of dynamically moving user traffic between different content delivery networks rather than relying on just one provider.

CDN visibility is the ability to collect, store, and analyze detailed event data from your content delivery networks.

Cold data is rarely accessed information that organizations keep for years to support compliance, security, and historical analysis. Understanding how cold tiers work, where they fall short, and where they are evolving helps teams balance cost against access without sacrificing visibility.

Column indexes are additional data structures in databases that include information about the location of data, making lookups much more efficient.

Data federation allows you to query multiple data sources as if they were one unified source.

A data lake is a centralized repository that stores vast amounts of structured, semi-structured, and unstructured data in its native formats or raw form.

Data retention controls how long your organization keeps data in accessible storage before archiving or deleting it.

Data tiering is the practice of distributing data across multiple storage layers, each with a different cost, performance, and retention profile, based on

An edge server is a computer that sits physically close to the people or devices using an application.

Federated search lets a single query interface reach across multiple, heterogeneous data sources simultaneously, collect partial results from each one.

Hot data is the information your systems need to read or write right now, supporting live dashboards, alerting, and analytics where seconds matter.

Hot data retention is the policy that sets how long recent, frequently queried data stays instantly accessible before it ages out or moves to cheaper, slower storage.

When data is immutable, once it’s created or written, it can’t be changed or modified.

When your analytics dashboard goes dark during a product launch or your security monitoring stops working during a suspected breach, you’re dealing with an incident.

Late-arriving data consists of log records or events that reach your data processing system after the time window they belong to has already closed.

When your application throws an error at 3 AM, log analytics is what helps you understand why. Every system in your infrastructure generates a continuous stream of events: application logs, database queries, network requests, security events, and performance metrics. Log analytics is the practice of collecting, storing, and analyzing these machine-generated events to support operations, security monitoring, and troubleshooting. Think of logs as your system’s memory. Just as you might review your calendar to underst…

Log management is the practice of collecting, storing, and analyzing the records your applications and servers produce. Every time a user clicks a button, a database runs a query, or a firewall blocks a connection, a log entry is created. On their own, these lines of text are small and unremarkable.

MELT stands for Metrics, Events, Logs, and Traces — the four signal types that modern software systems emit to support monitoring, debugging, and performance analysis.

Object storage is a data storage architecture that manages and stores data as objects that contain the data itself as well as metadata and a unique identifier. It is typically decoupled and distributed.

Observability allows you to measure the internal state of a system (such as applications, infrastructure, and services) by analyzing its external outputs.

Observability 2.0 describes a fundamental rethink of how engineering teams collect, store, and query telemetry data.

OLAP allows users to analyze data in multiple dimensions for fast decision-making. OLAP solutions power a wide range of use cases from log analytics to business intelligence.

OpenTelemetry (OTel) is a vendor-neutral, open-source observability framework hosted by the Cloud Native Computing Foundation (CNCF).

Real-time analytics processes data within seconds of generation to enable immediate decisions.

Real-time data refers to information that flows continuously from sources like applications, devices, or content delivery networks and gets processed quickly enough to support immediate decision-making. The “real-time” designation depends entirely on your use case. Google’s ad bidding system completes its entire decision loop in around 100 milliseconds, while Cloudflare’s operational logs typically arrive within two seconds for troubleshooting purposes. Organizations care about real-time data because it enables the…

A security data lake is a centralized, scalable repository that ingests, stores, and makes queryable large volumes of security-relevant telemetry.

Security logs are records of security-relevant events generated by every piece of technology in your environment.

Stateless architecture treats every request as independent and pushes persistent state to external stores, letting compute scale, recover, and replace itself without losing user data.

Stream processing is the real-time processing of continuous data streams.

Summary tables, also known as aggregate tables, store aggregations of raw data, providing fast queries for aggregations and metrics.

Telemetry is the automated process of collecting, transmitting, and analyzing data from sources to monitor and control systems.

A web application firewall (WAF) is acts as a protective barrier for web applications and APIs. WAFs inspect incoming HTTP traffic and filter or block unwanted traffic, such as malicious requests.

A wide event is a single structured record that captures every known attribute about one unit of work — one request, one transaction, one span — in a