Amazon EKS — Glossary
3 minute read
Key terms and definitions used throughout this Amazon EKS tutorial.
A
| Term |
Definition |
| Access Entry |
EKS-native API for granting IAM principals access to a cluster (replacement for aws-auth ConfigMap) |
| Add-on |
Pre-configured operational software managed by EKS (e.g., VPC CNI, CoreDNS, kube-proxy) |
| ALB |
Application Load Balancer — Layer 7 load balancer for HTTP/HTTPS traffic |
| ArgoCD |
GitOps continuous delivery tool for Kubernetes |
| aws-auth ConfigMap |
Legacy method of mapping IAM identities to Kubernetes RBAC (in kube-system namespace) |
C
| Term |
Definition |
| Cluster Autoscaler |
Kubernetes add-on that adjusts node group size based on pending pods |
| ConfigMap |
Kubernetes object for storing non-sensitive configuration data |
| Control Plane |
The managed EKS components (API server, etcd, scheduler, controller manager) — AWS manages these |
| CoreDNS |
DNS server for Kubernetes service discovery — runs as a Deployment in kube-system |
| CrashLoopBackOff |
Pod status indicating the container keeps crashing and Kubernetes is backing off restarts |
D
| Term |
Definition |
| DaemonSet |
Ensures a copy of a pod runs on every (or selected) node |
| Deployment |
Kubernetes resource for managing stateless application replicas |
E
| Term |
Definition |
| EBS CSI Driver |
Add-on enabling Kubernetes to use Amazon EBS for persistent storage |
| ECR |
Elastic Container Registry — AWS managed Docker image registry |
| EKS |
Elastic Kubernetes Service — AWS managed Kubernetes |
| ENI |
Elastic Network Interface — each pod IP comes from an ENI in the VPC (via VPC CNI) |
| eksctl |
Official CLI tool for creating and managing EKS clusters |
F
| Term |
Definition |
| Fargate |
Serverless compute for EKS — AWS manages the nodes |
| Fargate Profile |
Defines which pods run on Fargate based on namespace and label selectors |
| Fluent Bit |
Lightweight log processor/forwarder — commonly used for container log shipping |
| Flux |
GitOps toolkit for Kubernetes — alternative to ArgoCD |
H
| Term |
Definition |
| Helm |
Package manager for Kubernetes — uses charts to define, install, and upgrade applications |
| HPA |
Horizontal Pod Autoscaler — scales pod replicas based on metrics |
I
| Term |
Definition |
| IAM |
Identity and Access Management — AWS authentication and authorization service |
| Ingress |
Kubernetes resource defining external access to Services (usually via ALB) |
| IRSA |
IAM Roles for Service Accounts — maps Kubernetes service accounts to IAM roles for AWS API access |
K
| Term |
Definition |
| Karpenter |
Node provisioner that launches right-sized compute in response to pending pods |
| Kubecost |
Cost monitoring tool for Kubernetes workloads |
| kubectl |
Command-line tool for interacting with Kubernetes API |
| kubelet |
Agent running on each node that manages pod lifecycle |
| kube-proxy |
Network component on each node that implements Kubernetes Service networking |
L
| Term |
Definition |
| LimitRange |
Sets default and maximum resource limits per namespace |
M
| Term |
Definition |
| Managed Node Group |
EKS-managed EC2 instances with automated provisioning, updates, and lifecycle |
N
| Term |
Definition |
| Namespace |
Virtual cluster partition for resource isolation and RBAC |
| NAT Gateway |
AWS service enabling outbound internet access from private subnets |
| NetworkPolicy |
Kubernetes resource controlling pod-to-pod network traffic |
| NLB |
Network Load Balancer — Layer 4 load balancer for TCP/UDP traffic |
| Node |
Worker machine (EC2 instance) running pods |
O
| Term |
Definition |
| OIDC |
OpenID Connect — identity protocol used by EKS for IRSA |
| OOMKilled |
Container terminated because it exceeded its memory limit (exit code 137) |
| OpenCost |
Open-source Kubernetes cost monitoring (CNCF project) |
P
| Term |
Definition |
| PDB |
Pod Disruption Budget — limits voluntary disruptions to maintain availability |
| PersistentVolume (PV) |
Cluster-level storage resource backed by EBS, EFS, or other storage |
| PersistentVolumeClaim (PVC) |
Pod’s request for persistent storage |
| Pod |
Smallest deployable unit in Kubernetes — one or more containers sharing network and storage |
| Prefix Delegation |
VPC CNI feature assigning /28 prefixes instead of individual IPs, enabling more pods per node |
R
| Term |
Definition |
| RBAC |
Role-Based Access Control — Kubernetes authorization model using Roles and RoleBindings |
| ResourceQuota |
Limits total resource consumption per namespace |
S
| Term |
Definition |
| Secret |
Kubernetes object for storing sensitive data (passwords, tokens) |
| SecurityContext |
Pod/container-level security settings (runAsUser, capabilities, readOnlyRootFilesystem) |
| Service |
Kubernetes abstraction providing stable network endpoint for pods (ClusterIP, NodePort, LoadBalancer) |
| ServiceAccount |
Kubernetes identity for pods to authenticate to the API server and AWS (via IRSA) |
| StatefulSet |
Kubernetes resource for managing stateful applications with stable identity and persistent storage |
| StorageClass |
Defines how persistent volumes are dynamically provisioned (gp3, io2, efs) |
T
| Term |
Definition |
| Taint |
Node attribute that repels pods unless they have a matching toleration |
| Terraform |
Infrastructure as Code tool by HashiCorp — commonly used for EKS cluster provisioning |
| Toleration |
Pod attribute allowing scheduling on tainted nodes |
| TopologySpreadConstraints |
Rules for distributing pods across topology domains (AZs, nodes) |
V
| Term |
Definition |
| VPA |
Vertical Pod Autoscaler — adjusts pod CPU/memory requests based on historical usage |
| VPC |
Virtual Private Cloud — isolated AWS network for your resources |
| VPC CNI |
AWS EKS networking plugin that assigns VPC IP addresses directly to pods |