Skip to content
FluidCloudBook a demo
CUSTOMER STORY / NASIKO

Nasiko builds a disaster recovery region for its EKS platform with FluidCloud

One cluster and 79 dependent resources, deployed from us-east-1 to us-west-2 and checked in one session.

Nasiko’s core Kubernetes platform ran in one AWS region. FluidCloud scanned the account and assessed all the dependencies of the cluster. It deployed the cluster into us-west-2 with its network, security groups and encryption. Then it confirmed that the copy matched the source.

Nasiko
Use case
Disaster recovery, cross-region
Cloud
AWS: one account, three regions onboarded
Workload
Amazon EKS cluster nasikoaws, Kubernetes 1.31
Source region
us-east-1 (N. Virginia)
DR region
us-west-2 (Oregon)
FluidCloud operations
Scan, Assess, Deploy, Recover
Completed
October 2026
AT A GLANCE
2 regions
Now run the nasikoaws control plane. Before, one region did.
79
Resources deployed into us-west-2: EKS, network, security and encryption
2,820
Resources in continuous, agentless scans across three regions
~190
AWS service categories in each daily snapshot

Figures from the FluidCloud scans of the Nasiko AWS account, 29–30 October 2026.

01

A single-region platform with no DR site

The Nasiko EKS control plane was only in us-east-1. A regional outage would stop the platform, with no disaster recovery site to fail over to.

A manual rebuild of an EKS cluster in another region is slow and risky. The cluster is one resource, but it depends on many more: a VPC, subnets across availability zones, route tables, and internet and NAT egress. It also depends on security groups and their rules, network ACLs, a KMS key for secrets, and the access mappings that control who can reach the API.

If one dependency is missing, the cluster does not create, or it starts but nobody can reach it. Nasiko had to see all that the cluster depends on, make all of it again in a second region, and prove that the copy matched.

Resiliency starts with a recovery region that you built and checked before the outage, not during it.

02

Scan, assess, deploy, recover

FluidCloud worked from the real state of the Nasiko account. It installed nothing and changed nothing in us-east-1.

us-east-1 · Source

EKS cluster nasikoaws

Kubernetes 1.31 · private and public API endpoints

  • NetworkingVPC, subnets, routes, NAT and internet egress
  • SecuritySecurity groups, rules, network ACLs
  • EncryptionRegional KMS key for Kubernetes secrets
  • Add-ons and access5 managed add-ons, access entries, logging

Deploy

us-west-2 · DR region

EKS cluster nasikoaws

Kubernetes 1.31 · the same configuration

  • 31NetworkingNew VPC, 6 subnets
  • 31Security4 groups, 16 rules, ACL
  • 3EncryptionNew regional KMS key
  • 14EKS and accessCluster, add-ons, logs

IAM, account-wide. The same cluster role and identities serve the two regions, so there are no duplicate roles to keep the same.

The counts in us-west-2 are the resources added between the FluidCloud scans of that region before and after the deployment. FluidCloud read the source cluster in us-east-1 and did not change it.
  1. 01

    Scan

    Nasiko connected its AWS account to FluidCloud with read-only access in us-east-1, us-west-1 and us-west-2. Each scan records about 190 AWS service categories, with no agent, and FluidCloud keeps seven daily snapshots for each region. The us-east-1 region alone has 1,129 resources, which include four EKS clusters.

  2. 02

    Assess

    FluidCloud mapped each resource and the relations between them. From nasikoaws, it found all the dependencies of the cluster: its VPC, subnets, routes, gateways, security groups, network ACLs, KMS key, add-ons and access entries. That set became the scope of the deployment, with nothing missing and nothing extra.

  3. 03

    Deploy into us-west-2

    FluidCloud generated Terraform for us-west-2 from the assessment and deployed it. It made new region-specific resources, such as a new VPC, new subnets and a new KMS key. The cluster kept its name, Kubernetes version, add-ons and access model.

  4. 04

    Recover

    Automatic scans of us-west-2 after the deployment recorded each new resource, and FluidCloud compared their configuration with the source cluster. The scans recorded the full deployment between 00:09 and 01:33 UTC on 30 October 2026.

03

What FluidCloud deployed in us-west-2

79 resources: the control plane and all that it needs to run in a new region.

IAM stays shared. IAM is account-wide in AWS, so the recovery cluster uses the same cluster role as the source and needs no duplicate roles. EKS access entries bind the service-linked role, the node role and the administrator to the new cluster.

The DR standby stays small. Nasiko deployed the control plane and its dependencies without worker node groups. It can attach compute in us-west-2 when it needs it.

04

Results

The FluidCloud scans of the two regions show the same cluster configuration. The region-specific resources are new.

Attributeus-east-1 sourceus-west-2 DR
Kubernetes version1.311.31
Managed add-onsvpc-cni, coredns, kube-proxy, aws-ebs-csi-driver, eks-pod-identity-agentThe same five add-ons
Access entries3, API and ConfigMap authentication3, API and ConfigMap authentication
Control-plane loggingAPI, audit, authenticatorAPI, audit, authenticator
API endpoint accessPrivate and publicPrivate and public
Control-plane subnets33, in a new VPC
Kubernetes service range172.20.0.0/16172.20.0.0/16
Secrets encryptionRegional KMS keyNew regional KMS key
Cluster IAM roleAccount-wide roleThe same role, shared

Source: FluidCloud configuration snapshots of the nasikoaws cluster in each region, 29–30 October 2026.

05

What is next

Daily snapshots keep the inventory current. Thus Nasiko can do the same deployment again when the source changes, add worker capacity in us-west-2 for a recovery drill, or use the same method for its other clusters. The same workflow moves workloads across AWS, Azure, GCP, OCI, VMware, Nutanix, OVHcloud and Vultr.

Build your disaster recovery regionwith FluidCloud

See FluidCloud scan your estate, assess the dependencies of a workload, and deploy it into a second region.

All customer stories