ACE-Professional key terms

A geographic area containing multiple isolated data center facilities, chosen so applications can run close to users while meeting latency, residency, and availability requirements.

An isolated location within a region with independent power, cooling, and networking, used to spread workloads so a single facility failure does not take down an application.

The ability of a system to keep operating correctly when a component fails, typically by duplicating resources so a standby takes over automatically without user-visible interruption.

A design approach that minimizes downtime by removing single points of failure and enabling quick recovery, usually expressed as a target uptime percentage such as 99.9 percent.

The ability to automatically add or remove resources as demand changes, so capacity tracks workload needs without manual intervention or long-term over-provisioning.

Adding or removing whole instances of a resource, such as more web servers behind a load balancer, to handle changing traffic levels.

Increasing or decreasing the size or power of a single resource, such as moving a database to a larger instance type, which often requires a restart.

Running an application in two or more geographic regions so it survives a full regional outage and can serve users from a closer location.

The set of plans and resources used to restore applications and data after a major failure, measured by recovery time and recovery point objectives.

The maximum acceptable elapsed time to restore a service after an outage, which drives how much standby infrastructure a recovery plan must keep ready.

The maximum acceptable amount of data loss measured in time, indicating how recent a backup or replica must be for recovery to be acceptable.

A software-defined computer with its own operating system running on shared physical hardware, giving full control over the guest environment.

A packaged unit of an application and its dependencies that shares the host operating system kernel, starting quickly and using fewer resources than a virtual machine.

A model where the provider manages servers entirely and runs code only when triggered, billing for execution time and requests rather than idle capacity.

A named hardware profile specifying the amount of virtual CPU, memory, and sometimes local storage or accelerators allocated to a compute instance.

A policy-driven service that adjusts the number of running compute instances based on metrics such as CPU utilization or request count to match demand.

A platform that schedules containers across a cluster, restarts failed ones, and handles service discovery, scaling, and rolling updates automatically.

A design where producers emit events and consumers react to them asynchronously, decoupling components so each can scale and fail independently.

A durable buffer that holds messages between producers and consumers, smoothing traffic spikes and letting consumers process work at their own pace.

A separate queue that receives messages a consumer repeatedly fails to process, so failures can be inspected and retried without blocking normal traffic.

A flat storage service for arbitrary files addressed by a key within a bucket, offering high durability and scale but no in-place file editing.

Storage presented to a compute instance as raw virtual disks, used for operating systems and databases that need low-latency read and write access.

A tier label on stored objects that balances retrieval cost, access latency, and durability, letting frequently and rarely accessed data use different pricing.

A rule set that automatically transitions objects to cheaper storage tiers or deletes them after a defined age, reducing cost without manual cleanup.

A point-in-time copy of a volume or database used for backup and cloning, typically stored incrementally and restorable to a new resource.

A resource-based access policy attached to an object storage bucket that states which principals may perform which actions on which keys.

A logically isolated network in the cloud where you define subnets, route tables, and gateways to control how resources communicate internally and externally.

A range of IP addresses within a virtual network, placed in one availability zone and marked public or private to control routing to the internet.

A stateful virtual firewall attached to individual resources that permits only explicitly allowed inbound and outbound traffic, evaluated by allow rules only.

A stateless subnet-level filter that evaluates ordered allow and deny rules on inbound and outbound traffic, applied to everything in the subnet.

A set of rules that determines where network traffic from a subnet is directed, such as to a local network, internet gateway, or private link.

A service that distributes incoming requests across multiple healthy backends, performs health checks, and improves availability and throughput.

A globally distributed cache that serves static and dynamic content from edge locations near users, reducing latency and origin load.

The system of users, groups, roles, and policies that determines who can perform which actions on which cloud resources.

A security principle of granting only the permissions needed for a task and nothing more, limiting the blast radius of mistakes or compromised credentials.

A set of permissions that can be assumed by a trusted principal or service, providing temporary credentials instead of long-lived access keys.

A structured document that allows or denies specific actions on specific resources, attached to identities or resources to define access.

Protecting stored data by encrypting it on disk so it is unreadable without the correct key, even if physical media or backups are exposed.

Protecting data while it moves between systems using protocols such as TLS, preventing interception or tampering on the network.

A managed service that creates, stores, rotates, and audits cryptographic keys, integrating with storage and database services for encryption.

A periodic probe that tests whether a resource is responding correctly, allowing load balancers and orchestrators to route traffic away from unhealthy targets.

A contractual commitment from a provider about availability or performance, usually with credits if the measured service level is not met.

The ability to understand a system's internal state from its outputs, combining metrics, logs, and traces to diagnose issues quickly.

A numeric time-series measurement such as CPU utilization or request latency, used for dashboards, alarms, and scaling decisions.

A technique that follows a single request across services by propagating a shared identifier, revealing where latency and errors originate.

Collecting logs from many sources into a central searchable store, enabling correlation and retention beyond the life of individual instances.

Rules that trigger notifications when a metric crosses a threshold or an event occurs, so operators respond before users notice problems.

Attaching key-value metadata to resources so costs, ownership, and environments can be grouped, filtered, and reported accurately.

The practice of attributing cloud spend to teams, projects, or products, usually through consistent tags and account structure, to enable accountability.

A spending threshold configured for an account or tag scope that sends alerts or triggers actions when forecast or actual cost approaches the limit.

A commitment to use a certain amount of compute for a one- or three-year term in exchange for a discounted rate compared with on-demand pricing.