DoubleTick builds a disaster recovery region for its core edge services
Its website, white-label partner proxy and NAT egress instance, with three VPCs, peering, a VPN gateway and all security rules: 80 resources, deployed from Mumbai to Oregon.
DoubleTick runs its production platform in AWS ap-south-1. FluidCloud scanned all 21,153 resources in its production account. It assessed the dependencies of three core instances and generated deployment-ready Terraform for us-west-2. The DoubleTick team applied it in its own environment. All three services now also run in a second region, with their full network and security footprint.
- Use case
- Disaster recovery, cross-region
- Cloud
- AWS: one production account, four regions onboarded
- Workloads
- Company website, white-label partner proxy and NAT egress instance
- Source region
- ap-south-1 (Mumbai)
- DR region
- us-west-2 (Oregon)
- FluidCloud operations
- Scan, Assess, Deploy, Recover
- Completed
- November 2026
- 21,153
- Resources scanned across four AWS regions, 12,335 of them in ap-south-1
- 80
- Resources deployed in us-west-2: compute, network and security
- 3
- Core production instances recovered into a second region
- 3 VPCs
- Rebuilt with their peering, VPN gateway and routes
Figures from the FluidCloud scans and deployments for the DoubleTick production AWS account, 2 November 2026.
A WhatsApp Business platform on AWS in Mumbai
DoubleTick, by QuickSell, is a WhatsApp Business platform. Companies use it for marketing, sales and support conversations on WhatsApp, with a shared team inbox, broadcast campaigns and chatbots. Its production platform runs on AWS in Mumbai.
This story is about three services at the edge of that platform: the public website, the proxy for the custom domains of white-label partners, and the NAT instance. The NAT instance gives DoubleTick a stable egress address that partners can allowlist.
Edge services that depend on many networks
The DoubleTick production estate has more than 12,000 resources in Mumbai. A regional outage would leave its edge services with no disaster recovery site.
The website, the partner proxy and the NAT instance use three VPCs. Peering connections join the VPCs, and one VPC has a VPN gateway. Traffic goes through internet gateways, an IPv6 egress gateway, a managed NAT gateway and an S3 endpoint. Security groups with more than 30 rules control it.
To copy this by hand into another region, the team must find each dependency again. One missing route or rule can make a service unreachable. DoubleTick wanted the full footprint as code, changed for a new region, and applied under its own control.
To recover a service, you must recover the networks that it runs on, not only the instance.
Scan, assess, deploy, recover
FluidCloud used read-only access to the production account. It installed nothing and changed nothing in ap-south-1.
ap-south-1 · Source
3 core production instances
Website · partner proxy · NAT egress
- ComputeInstances, machine image, key pairs, interfaces
- Networking3 VPCs, peering, VPN, NAT, gateways
- SecuritySecurity groups with ingress and egress rules, default network ACL
Deploy
us-west-2 · DR region
3 core instances, running
80 resources deployed
- 9Compute3 instances, image copy
- 35Networking3 VPCs, 7 subnets, VPN
- 36Security4 security groups, 31 rules, default network ACL
IAM, account-wide. The partner proxy and the NAT instance use their instance profiles as they are. FluidCloud copied no roles.
- 01
Scan
DoubleTick connected its production AWS account to FluidCloud with read-only access in ap-south-1, us-east-1, us-west-1 and us-west-2. Daily snapshots record the configuration of all 21,153 resources, with no agent: 12,335 in ap-south-1, 3,608 in us-west-2, 2,937 in us-east-1 and 2,273 in us-west-1.
- 02
Assess
FluidCloud started from each instance and found all that it uses: subnets and route tables, the three VPCs with their peering and VPN connections, and each gateway to the internet. It also found the security groups and their rules, and the IAM instance profiles of the proxy and the NAT instance. The website and the partner proxy are two workloads on one shared network.
- 03
Deploy
FluidCloud generated Terraform for us-west-2, one bundle for each workload. It copied the machine image of the website into the new region, and rebuilt the region-specific resources, such as subnets, gateways and endpoints, for Oregon. The DoubleTick team applied the two bundles in its own environment. The team keeps the credentials and the state.
- 04
Recover
The website, the white-label partner proxy and the NAT egress instance now run in us-west-2, on a full copy of their network and security footprint. DoubleTick has a disaster recovery region for its edge services.
What FluidCloud deployed in us-west-2
80 resources: compute, network and security.
- Compute · 9
Three instances and their image
The three core instances, a copy of the website machine image in us-west-2, two key pairs and three network interfaces.
- Networking · 35
Three VPCs and their connections
Three VPCs with seven subnets, which include the EKS control-plane and internal-apps subnets. Three internet gateways, an IPv6 egress-only gateway, a NAT gateway and three Elastic IPs. Three route tables with four associations, two VPC peering connections with their accepters, a VPN gateway and its attachment, DHCP options with two associations, and an S3 gateway endpoint.
- Security · 36
Groups, rules and ACL
Security groups for the website, the partner proxy and the NAT instance, and the default group of the VPC, with 23 ingress and 8 egress rules. The default network ACL.
- Identity · shared
IAM used again, not copied
The partner proxy and the NAT instance use the same instance profiles as their sources. IAM is account-wide in AWS, so the two regions use one identity model.
Two workloads, one shared foundation
Each workload has its own deployment bundle, and the two use the same networks. Each resource counted one time, the total is 80.
| Deployment | What it includes | Resources |
|---|---|---|
| Website | The website instance, its image copy, key pair, interface and security group, and the shared foundation | 69 |
| Partner proxy | The partner proxy instance, its Elastic IP, key pair, interface and security group, and the shared foundation | 52 |
| Shared by the two | Three VPCs, seven subnets, gateways, NAT, routes, peering, VPN, the NAT instance and common rules | −41 |
| Unique resources in us-west-2 | 80 |
Results
Each of the three core instances kept its size, storage and access settings in us-west-2.
| Instance | What it does | In us-west-2 |
|---|---|---|
| doubletick-website | Serves the DoubleTick company website on Node.js behind nginx | t4g.medium (Arm), 8 GiB gp3, started from the website image copied into the region, token-only instance metadata |
| wl-partner-proxy | Serves the custom domains of white-label partners | t3.small, 15 GiB gp3 encrypted, its own Elastic IP, the existing instance profile, token-only instance metadata |
| fck-nat | Gives private workloads a stable egress address that partners can allowlist | c8gn.medium (Arm), fck-nat image, 4 GiB gp3, the existing instance profile, token-only instance metadata |
Source: the FluidCloud “doubletick-website-DR” and “PartnerAWS to West2 DR” deployments for the DoubleTick production account, 2 November 2026.
What is next
Daily snapshots keep the inventory current. Thus DoubleTick can generate each bundle again when the source changes, practise recovery into us-west-2 on a schedule, and add more of its 12,000-resource Mumbai estate. The same workflow moves workloads across AWS, Azure, GCP, OCI, VMware, Nutanix, OVHcloud and Vultr.
Recover your services withthe networks they run on
See FluidCloud scan your estate, assess all that a workload depends on, and deploy it into a second region.