geoffsdesk
- 14 skills
- 0 followers
- 6 hours ago last updated
- ▌ Eks Discovery · geoffsdeskInventory an Amazon EKS estate end-to-end. Enumerates clusters, node groups, workloads, addons, IAM (IRSA), networking (ALBs, NLBs, security groups), storage (PVs, EBS, EFS), data dependencies (RDS, ElastiCache), and observability config. Use when starting an EKS-to-GKE migration, when "we need to know what's actually in this account", when producing a pre-migration inventory, or when the user says "audit our EKS environment". Outputs a structured inventory.json plus a human-readable summary.
- ▌ Data Migration · geoffsdesk bundlePlan and (with confirmation) execute data migrations from AWS data services to GCP equivalents — RDS to Cloud SQL or AlloyDB, ElastiCache to Memorystore, S3 to GCS, MSK to GCP Kafka or Confluent Cloud — and the secrets/config moves that go with them. Produces per-data-system runbooks, replication plans, validation gates, and cutover scripts. Use after storage-translation, when "plan our RDS migration", "migrate Redis to Memorystore", or "data layer plan for the GKE move".
- ▌ Traffic Cutover · geoffsdeskShift production traffic from EKS to GKE per workload with weighted DNS, gateway-level traffic split, or service-mesh routing — including pre-cutover validation, gradual ramp (1% → 10% → 50% → 100%), SLO-tied auto-rollback gates, and post-cutover soak. Produces a per-workload cutover runbook with explicit go/no-go gates. Use when "plan the traffic cutover", "shift traffic gradually to GKE", or "execute the move for service X".
- ▌
- ▌ Rollback Playbook · geoffsdeskRoll back a workload from GKE to EKS when a cutover gate breaches or when a regression is found. Reverses traffic shifts, repoints data writers if applicable, restores config, captures evidence, and produces a postmortem-ready timeline. Use when a traffic-cutover gate trips, when "rollback service X to EKS", or when an incident requires returning traffic to the source.
- ▌ Post Migration Ops · geoffsdeskRun the day-2 work after every workload is on GKE — FinOps tuning (right-sizing, autoscaling, CUDs, Spot adoption), security hardening (Workload Identity audit, NetworkPolicy enforcement, Binary Authorization mode flip), observability completeness (golden-signal dashboards, SLO-tied alerts), and EKS decommissioning. Produces a final summary, FinOps report, hardening checklist, and decommission plan. Use as Phase 5 of a Portage migration, when "what do we do post-migration", or "tune GKE cost / harden the new environment".
- ▌ Registry Migration · geoffsdeskMigrate container images from Amazon ECR to Google Artifact Registry. Plans the AR repo layout, mirrors images (skopeo or pull-through), updates build pipelines, sets up vulnerability scanning, and produces an image-map.json mapping every old image reference to its new one. Use when "migrate ECR to Artifact Registry", "set up image registries on GCP", or as Phase 3 of a Portage migration.
- ▌ Network Translation · geoffsdeskTranslate EKS network constructs to GKE equivalents. Maps ALBs/NLBs to GKE Gateways and Services, AWS Load Balancer Controller annotations to Gateway/Service annotations, Route 53 records to Cloud DNS, security groups to VPC firewall rules and NetworkPolicy, and AWS WAF rules to Cloud Armor policies. Produces translated manifests plus a network design diff. Use after gke-landing-zone, when "translate our ingress to GKE", "convert ALB ingresses to Gateway API", or "what's the network plan for the migration".
- ▌
- ▌ Identity Translation · geoffsdeskTranslate EKS IRSA (IAM Roles for Service Accounts) bindings to GKE Workload Identity. Maps every annotated ServiceAccount to a Google Service Account, translates AWS IAM policies into the closest GCP IAM role set, produces additive bindings on both sides for co-existence, and emits a verification plan. Use after gke-landing-zone, when "translate IRSA to Workload Identity", "set up identity for the GKE migration", or "what's the identity plan".
- ▌ Migration Assessment · geoffsdeskScore an EKS estate's readiness for migration to GKE and produce a graded readiness report with named blockers, risk ratings per workload, a phased migration plan, and a costed effort estimate. Consumes the inventory.json produced by eks-discovery. Use after discovery to answer "is this ready to migrate?", "what will it take?", "what blocks us?", "what's our migration plan?".
- ▌ Portage Orchestrator · geoffsdeskRun an end-to-end EKS to GKE migration program. Coordinates discovery, design, translation, cutover, and post-migration ops by sequencing the other Portage skills. Use when the user asks to "migrate from EKS to GKE", "plan an EKS→GKE migration", "replace our PSO engagement", or any holistic move of a Kubernetes estate from AWS to Google Cloud. Do NOT use for single-skill tasks (e.g., "translate this Helm chart") — call the specific skill directly.
- ▌ Workload Translation · geoffsdeskTranslate Kubernetes manifests, Helm charts, and Kustomize overlays from EKS conventions to GKE conventions. Rewrites AWS-specific annotations, replaces ALB Ingress with Gateway API, swaps ebs.csi.aws.com StorageClasses to pd.csi.storage.gke.io, fixes node selectors and tolerations, removes IRSA annotations in favor of Workload Identity, and produces a clean diff per workload. Use when "translate this Helm chart for GKE", "convert manifests to GKE", or after identity-translation in a Portage migration.
- ▌ Observability Translation · geoffsdeskTranslate EKS observability stacks to GKE — CloudWatch Logs and Metrics to Cloud Logging and Cloud Monitoring, Managed Prometheus (AMP) to Google Managed Service for Prometheus (GMP), CloudWatch Container Insights to GKE-native dashboards, X-Ray to Cloud Trace, and CloudWatch alarms to alerting policies. Produces translated alerting rules, dashboard JSON, log routing, and SLO definitions. Use when "translate observability", "set up monitoring on GKE", or as part of Phase 3 of a Portage migration.