SYSTEM Cited by 1 source

Generali Malaysia EKS platform¶

Generali Malaysia's EKS platform is the multi-tenant, AWS-managed EKS cluster (operated in EKS Auto Mode — AWS-managed K8s data plane on Bottlerocket nodes) that hosts Generali Malaysia's digital applications and several core insurance systems. Migration to AWS began 2019; EKS Auto Mode adoption is ongoing as of the 2026-03 source. This page synthesises the integration topology across six AWS services documented by the AWS Architecture Blog post.

Integration topology¶

                                                          Amazon
                                                          Managed
                                                          Grafana
                                                          (per-ns
                                                           dashboards)
                                                             ▲
                                                             │ CloudWatch
                                                             │ data source
 Users ──► Route 53 / ALB ──► Network Firewall (public subnet)
                                 │ SNI allow-list           │
                                 ▼                          │
                               NAT GW (protected subnet)    │
                                 │                          │
                         ┌───────┴──────────────────────────┴─┐
                         │                                    │
                         ▼                                    │
                ┌── EKS Auto Mode cluster ──┐                 │
                │  (Bottlerocket nodes,     │                 │
                │   weekly AMI replacement,  │                │
                │   PDBs + NDBs,             │                │
                │   off-peak windows)        │                │
                │                            │                │
                │  Namespace A (tenant 1)    │                │
                │    Deployments (Helm)      │                │
                │    HPA auto-scaling        │                │
                │    Stateless-only pods     │                │
                │    ExtSecretsOperator      │ ◄── Secrets Manager
                │                            │                │
                │  Namespace B (tenant 2)    │                │
                │  ...                       │                │
                └────────────┬───────────────┘                │
                             │                                │
                             │                                │
                             ▼                                │
                    CloudWatch Logs / Metrics ────────────────┘
                    GuardDuty (EKS audit + runtime + CloudTrail)
                    Inspector (ECR → running containers)

 Cost: cluster-name / namespace / deployment / node cost-allocation
       tags → Cost Explorer (unified with non-EKS AWS spend)
 Savings Plans applied to compute

Well-Architected pillar mapping¶

The source organises the design around AWS's six Well-Architected pillars:

Operational Excellence — EKS Auto Mode handles node lifecycle, OS upgrades, add-on upgrades, cluster-version upgrades. Shared platform-style multi-tenant cluster with namespace-scoped observability.
Reliability — PDBs + NDBs + off-peak maintenance windows (patterns/disruption-budget-guarded-upgrades); HPA auto- scaling; stateless-only pods simplify recovery.
Performance Efficiency — Auto Mode picks instance sizes from a customer-defined pool; HPA scales pods on real traffic; pods treated as immutable.
Security — GuardDuty (EKS protection + runtime), Inspector (ECR vulns with running-pod context), Network Firewall (SNI egress allow-list), ESO + Secrets Manager (env-var secret injection), IAM integration.
Cost Optimization — EKS Auto Mode's right-sizing + AWS Billing split cost allocation data for EKS (cluster-name / deployment / namespace / node tags) + Savings Plans.
Sustainability — right-sizing via Auto Mode; minimising over-provisioned footprint.

Operating principles (stated by Generali)¶

Only stateless micro-services on EKS.
Pods treated as immutable — upgrade by replacement.
Helm charts as the standardised deployment mechanism across services.
HPA-based traffic-driven pod auto-scaling.

These four compound: stateless-only makes pod immutability cheap; pod immutability makes HPA safe; Helm encodes all three as template defaults; and the composition is what makes EKS Auto Mode's platform-driven node churn compatible with production workloads.

Named failure modes handled¶

Auto Mode weekly AMI replacement terminates all replicas at once → Pod Disruption Budgets.
Simultaneous bulk node replacement → Node Disruption Budgets.
Off-peak alignment → customer-configured maintenance window.
Stateful pod migration during node churn → avoided structurally by the stateless-only rule.

Reported outcomes (qualitative)¶

"Significant reduction in operational overhead with EKS Auto Mode."
"Enhanced security with automated threat detection and response."
"Reduction in infrastructure costs through optimization."
"Improved mean-time-to-resolution."
"Accelerated application deployment cycles."

No percentages, cluster sizes, pod counts, or cost deltas published.

Stated future scope¶

Hosting AI models on the same EKS platform.
Hosting upcoming "agentic applications" on the same EKS platform.

No architectural detail given for either.

Seen in¶

sources/2026-03-23-aws-generali-malaysia-eks-auto-mode — the sole source for this platform as of 2026-04-21. Tier-1 (AWS Architecture Blog) customer-case-study format: substance is in the integration topology, not in quantified outcomes.