Book 05 · Cheat sheet

DevOps & Cloud

Tools and practices for building, testing, packaging and shipping software to production.

01Ansible: Ansible is an open-source automation tool that configures servers and deploys applications by running YAML playbooks over SSH, with no agent on the machines.
Ansible automates server configuration and application deployment.
Tasks are written in YAML playbooks and run against an inventory of hosts.
It is agentless: it connects over SSH and needs only Python on the hosts.
02Autoscaling: Autoscaling is the automatic adding or removing of computing resources, such as servers or containers, based on demand to keep performance steady and costs low.
Autoscaling adds or removes capacity automatically based on demand.
It is driven by metrics such as CPU, memory, request rate, or queue length.
Minimum and maximum limits keep scaling safe and costs predictable.
03AWSAmazon Web Services: AWS (Amazon Web Services) is Amazon's cloud platform: more than 200 pay-as-you-go services for servers, storage, databases and much more.
AWS is Amazon's cloud platform, with more than 200 pay-as-you-go services.
Core services include EC2 (servers), S3 (storage), RDS (databases) and Lambda.
Regions contain several availability zones for resilience.
04AzureMicrosoft Azure: Microsoft Azure is Microsoft's cloud computing platform, offering hundreds of on-demand services such as virtual machines, databases and AI models worldwide.
Azure is Microsoft's public cloud platform, launched in 2010.
It is one of the three largest clouds, with AWS and Google Cloud.
Services range from VMs and Kubernetes to databases, functions and AI models.
05Blue-Green Deployment: Blue-green deployment is a release strategy that uses two identical production environments and moves all traffic from the old version to the new one at once.
Two identical environments exist: one live, one idle.
The new version is deployed and tested on the idle environment before going live.
Traffic switches all at once, typically at the load balancer, with no downtime.
06Canary Deployment: A canary deployment releases a new software version to a small share of users first, checks its health, and then gradually rolls it out to everyone.
A canary sends a small percentage of traffic to the new version first.
Traffic is increased in steps while metrics are watched.
A bad release affects only a few users and can be rolled back quickly.
07CDNContent Delivery Network: A CDN is a network of servers spread around the world that stores copies of website content and delivers it to each user from the nearest location.
A CDN serves content from servers close to each user.
It works by caching copies of files from the origin server.
CDNs reduce latency, lower origin load, and help absorb traffic spikes and attacks.
08Chaos Engineering: Chaos engineering is the practice of deliberately injecting failures into a system, such as crashing servers, to confirm that it keeps working as expected.
Chaos engineering injects real failures on purpose to reveal hidden weaknesses.
Each experiment starts from a measurable steady state and a clear hypothesis.
Experiments keep the blast radius small and can be aborted at any time.
09CI/CDContinuous Integration / Continuous Delivery: CI/CD is a set of automated practices that build, test, and release code changes frequently, so software can be delivered to users quickly and safely.
CI automatically builds and tests every change merged into the shared codebase.
CD keeps the code always ready to release, or releases it automatically.
Pipelines are defined in configuration files stored alongside the code.
10Cloud Computing: Cloud computing is the on-demand delivery of computing resources, such as servers, storage, and databases, over the internet with pay-as-you-go pricing.
Cloud computing rents servers, storage, and services over the internet on demand.
The three main service models are IaaS, PaaS, and SaaS.
Elasticity lets resources grow and shrink automatically with demand.
11Configuration Management: Configuration management is the practice of defining the desired state of servers and software in code and using tools to apply and keep it automatically.
Configuration management keeps the desired state of systems in version-controlled code.
Tools apply that state automatically and correct drift on existing machines.
Idempotency makes it safe to apply the same configuration many times.
12Container: A container is a lightweight, isolated package that bundles an application with its dependencies and runs it on the host's shared operating system kernel.
A container packages an app with its dependencies so it behaves the same everywhere.
Containers share the host operating system kernel, unlike virtual machines.
Linux namespaces provide isolation, and cgroups limit resources.
13Container Registry: A container registry is a storage and distribution service for container images, letting teams push built images and pull them onto any server that runs them.
A registry stores container images and serves them to the machines that run them.
Images are grouped into repositories and versioned with tags.
A digest identifies an image's exact contents and cannot be moved like a tag.
14DevOpsDevelopment and Operations: DevOps is a set of practices and a culture that brings software development and IT operations together to deliver software faster and more reliably.
DevOps unites development and operations into one shared responsibility.
Automation is central: CI/CD, infrastructure as code, and monitoring.
Small, frequent releases reduce risk and speed up feedback.
15Distributed Tracing: Distributed tracing is a technique that follows a single request as it travels through many services, recording how long each step took and where it failed.
A trace follows one request end to end across services.
Traces are made of spans, each measuring one operation, linked in parent-child order.
A trace ID is passed between services, usually in the traceparent header.
16DNSDomain Name System: DNS is the internet's naming system that translates human-readable domain names like example.com into the numeric IP addresses computers use to connect.
DNS translates domain names into IP addresses.
Resolvers find answers by querying root, top-level domain, and authoritative name servers.
Common record types are A, AAAA, CNAME, MX, and TXT.
17Docker: Docker is an open-source platform for packaging an application and everything it needs into a container that runs the same way on any machine.
Docker packages an application and its dependencies into a container image.
A Dockerfile defines how the image is built.
An image is the template; a container is a running instance of it.
18Docker Compose: Docker Compose is a tool for defining and running multi-container applications, such as a web server plus a database, from one YAML file with one command.
Docker Compose runs multi-container applications from one YAML file.
docker compose up starts every service on a shared network.
Services declare images, ports, environment variables, volumes and health checks.
19Edge Computing: Edge computing runs code and processes data close to where users or devices are, instead of in a distant central data center, to reduce latency.
Edge computing runs code close to users or devices to reduce latency and bandwidth use.
On the web, edge functions run in many locations, often on CDN networks.
In IoT, edge devices process data locally and send only results to the cloud.
20Environment Variable: An environment variable is a named value set outside a program, by the operating system or runtime, that the program reads to configure its behavior.
Environment variables are key-value settings provided by the environment, not the code.
They let the same code run with different configuration in each environment.
Child processes inherit a copy of their parent's environment variables.
21Feature Flag: A feature flag is a switch in code that turns a feature on or off at runtime, letting teams deploy code without releasing it to every user at once.
A feature flag turns functionality on or off at runtime without a new deployment.
It separates deploying code from releasing a feature to users.
Targeting rules can enable a feature for specific users, groups, or a percentage of traffic.
22GitHub Actions: GitHub Actions is GitHub's built-in automation platform: YAML workflows in a repo run tests, builds and deployments on events like a push or pull request.
GitHub Actions runs automated workflows inside GitHub.
Workflows are YAML files in .github/workflows, triggered by events.
Jobs run on GitHub-hosted or self-hosted runners, step by step.
23GitOps: GitOps is a way of managing infrastructure and deployments where Git holds the desired state of a system and an automated agent keeps the live system in sync.
Git is the single source of truth for the desired state of the system.
Changes go through commits and pull requests, not manual commands.
An agent continuously reconciles the live system with the repository.
24Google CloudGoogle Cloud Platform: Google Cloud is Google's public cloud platform, offering compute, storage, data and AI services on the global infrastructure behind Google's own products.
Google Cloud is Google's public cloud platform, often called GCP.
It started with App Engine in 2008.
Compute Engine, Cloud Run and GKE run code; Cloud Storage and Cloud SQL store data.
25Grafana: Grafana is an open-source tool for building dashboards that turn metrics, logs and traces from many data sources into live charts and alerts in one place.
Grafana builds dashboards and alerts from existing data sources.
It connects to Prometheus, Loki, Elasticsearch, SQL databases and more.
Panels from different sources can be combined on one dashboard.
26Helm: Helm is the package manager for Kubernetes: it bundles an app's configuration files into a chart you can install, upgrade and roll back with one command.
Helm is the package manager for Kubernetes.
A chart bundles templated manifests with default values in values.yaml.
Each install is a release that can be upgraded and rolled back.
27IaaSInfrastructure as a Service: IaaS (infrastructure as a service) is a cloud model where you rent virtual machines, storage and networks on demand and manage the OS and above yourself.
IaaS rents virtual machines, storage and networks on demand.
You manage the operating system, software, patches and scaling.
Amazon EC2, launched in 2006, defined the model.
28Immutable Infrastructure: Immutable infrastructure is an approach where servers are never changed after deployment; every update replaces them with new, freshly built ones.
Servers and containers are never modified after they are deployed.
Every change produces a new image, and old instances are replaced, not patched.
The same tested image moves unchanged through staging and production.
29Infrastructure as Code: Infrastructure as code is the practice of defining servers, networks, and other infrastructure in version-controlled files that tools apply automatically.
IaC defines infrastructure in code files instead of manual console clicks.
Declarative tools compare the desired state with reality and apply the difference.
Infrastructure changes are versioned, reviewed, and rolled back like application code.
30Jenkins: Jenkins is an open-source automation server that builds, tests and deploys software through pipelines, and one of the oldest and most widely used CI/CD tools.
Jenkins is a self-hosted, open-source automation server for CI/CD.
It forked from Hudson and took its name in 2011.
Pipelines are defined in a Jenkinsfile with build, test and deploy stages.
31kubectl: kubectl is the command-line tool for Kubernetes: it sends requests to a cluster's API server to deploy applications, inspect them and change them.
kubectl is the Kubernetes command-line tool; every command is a call to the API server.
get, describe, logs and exec inspect what is running.
kubectl apply -f makes the cluster match YAML files that describe the desired state.
32Kubernetes: Kubernetes is an open-source system that automates deploying, scaling, and managing containerized applications across a cluster of machines.
Kubernetes orchestrates containers across a cluster of machines.
You declare the desired state in YAML, and Kubernetes keeps it that way.
Pods, deployments, and services are its core building blocks.
33Linux: Linux is an open-source operating system kernel that powers most servers, cloud platforms, containers, and Android phones, usually packaged as a distribution.
Linux is an open-source kernel; distributions add tools to make a full operating system.
It runs most servers, cloud infrastructure, supercomputers, and Android devices.
Containers depend on Linux kernel features such as namespaces and cgroups.
34Load Balancer: A load balancer is a server or service that spreads incoming traffic across several backend servers so no single one is overloaded and the app stays available.
A load balancer distributes requests across multiple servers.
Health checks automatically take failing servers out of rotation.
Common algorithms include round robin, least connections, and IP hash.
35Logging: Logging is the practice of recording timestamped messages about events in a running program, such as errors and requests, so people can investigate them later.
Logging records timestamped events from a running program for later investigation.
Log levels such as DEBUG, INFO, WARN, and ERROR indicate severity and control volume.
Structured logging writes entries as JSON with named fields, which makes them easy to query.
36Metrics: Metrics are numeric measurements of a system collected over time, such as request rate, error rate and CPU usage, used for dashboards, alerts and planning.
Metrics are numeric measurements recorded over time and stored as time series.
The common types are counters, gauges, and histograms.
Percentiles such as p95 and p99 reveal slow requests that averages hide.
37Object Storage: Object storage is a way of storing data as whole objects, each with a unique key and metadata, in flat buckets that scale to huge numbers of files over HTTP.
Object storage saves data as objects: bytes plus metadata, addressed by a unique key.
Objects live in buckets with a flat namespace and are accessed over HTTP.
Objects are replaced as a whole rather than edited in place.
38Observability: Observability is the ability to understand what is happening inside a running software system by collecting and analyzing its logs, metrics, and traces.
Observability means understanding a system's internal state from the data it emits.
Logs, metrics, and traces are the three main types of telemetry.
Distributed tracing follows one request across many services.
39OpenTelemetry: OpenTelemetry is an open standard and set of tools for collecting traces, metrics and logs from software and sending them to any monitoring backend.
OpenTelemetry is a vendor-neutral standard for traces, metrics and logs.
Code is instrumented once, and the data can go to any backend.
OTLP is its protocol; the Collector receives, processes and forwards data.
40PaaSPlatform as a Service: PaaS (platform as a service) is a cloud model where you deploy your code and the provider runs everything under it: servers, operating systems and scaling.
PaaS runs your code while the provider manages servers and the OS.
Heroku popularized it; Vercel, Render and App Service are examples.
Platforms add databases, scaling, logs, previews and rollbacks.
41Pod: A pod is the smallest deployable unit in Kubernetes: one or more containers that share a network address and storage and are scheduled together on one node.
A pod is the smallest unit that Kubernetes schedules and manages.
Containers in the same pod share an IP address, localhost, and volumes.
Most pods run one container; helper containers in the same pod are called sidecars.
42Postmortem: A postmortem is a written review after an incident that explains what happened, why it happened, and what the team will change so it doesn't happen again.
A postmortem analyzes an incident after it is resolved and records the lessons learned.
It includes a timeline, impact, root causes, contributing factors, and action items.
Blameless postmortems focus on system and process failures, not on individuals.
43Prometheus: Prometheus is an open-source monitoring system that collects metrics from apps and servers, stores them as time series and alerts when values cross a limit.
Prometheus is an open-source monitoring system and time-series database.
It pulls metrics by scraping HTTP endpoints, usually /metrics.
Each series has a name and labels; PromQL queries them.
44Reverse Proxy: A reverse proxy is a server that sits in front of web servers, accepts client requests on their behalf, and forwards each request to the right backend server.
A reverse proxy receives requests on behalf of backend servers and forwards them.
It commonly handles HTTPS termination, caching, compression, and routing.
Backend servers can stay hidden on a private network.
45Rollback: A rollback is the process of returning software to a previous, known-good version after a new deployment causes errors, outages, or other unexpected problems.
A rollback restores the last known-good version of an application after a bad release.
It is usually the fastest way to reduce user impact during an incident.
Blue-green deployments and versioned container images make rollbacks quick.
46SaaSSoftware as a Service: SaaS (software as a service) is software delivered over the internet as a subscription; users sign in while the provider runs, updates and secures it.
SaaS is software used over the internet as a subscription.
The provider hosts, updates, secures and backs up the application.
It is the top layer above PaaS and IaaS, with the least to manage.
47Serverless: Serverless is a cloud model in which the provider runs your code on demand, manages all the servers, scales automatically, and bills only for actual use.
Serverless does not mean no servers; it means you don't manage them.
Code runs in response to events such as HTTP requests or file uploads.
It scales automatically, including down to zero when idle.
48Service Mesh: A service mesh is an infrastructure layer that manages traffic between microservices, adding encryption, retries, routing, and monitoring without code changes.
A service mesh manages service-to-service traffic in a microservices system.
Proxies form the data plane, and a control plane configures them.
It provides mutual TLS, retries, timeouts, traffic splitting, and telemetry.
49Site Reliability Engineering: Site reliability engineering is a discipline that applies software engineering to operations, keeping services reliable with automation and measurable targets.
SRE applies software engineering to operations and infrastructure work.
Reliability is measured with SLIs and SLOs that reflect what users experience.
An error budget balances the pace of new releases against stability.
50SLAService Level Agreement: An SLA (service level agreement) is a provider's commitment to customers about the level of service, such as 99.9% uptime, and what happens if it isn't met.
An SLA is a formal promise about service quality, such as uptime.
Missing it usually earns customers service credits.
99.9% allows about 43 minutes of downtime a month; 99.99% about 4.3 minutes.
51SLOService Level Objective: An SLO is a measurable reliability target for a service, such as 99.9% of requests succeeding over 30 days, that tells a team how reliable is reliable enough.
An SLO is a target value for a reliability measurement over a time window.
The measurement itself is called an SLI, for example the share of successful requests.
The allowed failure, 100% minus the SLO, is the error budget.
52Terraform: Terraform is an infrastructure-as-code tool: you describe cloud resources in configuration files, and one command creates or updates them to match.
Terraform describes cloud infrastructure as code, in HCL files.
terraform plan previews changes; terraform apply makes them.
A state file records which real resources Terraform manages.
53Virtual Machine: A virtual machine is a software-based computer that runs its own operating system on shared physical hardware, isolated from other machines on the same host.
A VM is a software-emulated computer with its own operating system.
A hypervisor shares one physical host's hardware among many isolated VMs.
A VM can run a different operating system than its host.
54YAMLYAML Ain't Markup Language: YAML is a human-readable data format that uses indentation instead of brackets, widely used for configuration files in DevOps tools and CI/CD pipelines.
YAML is a human-readable format for structured data and configuration.
Indentation with spaces defines structure; tabs are not allowed.
It supports comments with #, unlike JSON.