After ransomware, the latest Terraform snapshot may be the wrong one
Why an approved artifact can still be unsafe for a specific incident, and how compromise-time selection aligns infrastructure with a trusted data point.
Ransomware
Insights
Technical ideas written for recovery decisions.
These summaries are drawn from the supplied Dockfield source articles and architecture briefs. They separate infrastructure reproducibility, data recoverability and incident-time trust.
DR-as-Code: rebuilding the environment around the data
Why backup and workload recovery are not the same as reconstructing networking, security, compute and dependencies.
RDS and Aurora point-in-time recovery vs AWS Backup
Point-in-time recovery for operational precision, AWS Backup for governance, and DR-as-Code for the surrounding platform.
Reverse-engineering AWS into Terraform for legacy estates
From Boto3 discovery to dependency-safe Terraform, including ID-to-reference translation and VPC-first validation.
Custom AWS disaster recovery with manual failover
Same-account and multi-account recovery, scheduled capture, target deployment and controlled DNS switching.
ECS Amazon Linux 2 to Amazon Linux 2023
Progressive capacity-provider placement, canary weights and state decoupling through EFS or external storage.
Editorial principle: recovery claims need scope and evidence. A configuration interval is not automatically the application's data RPO, and an approved artifact is not automatically safe for every incident.
Free initial consultation
Have a technical recovery question?
Bring one workload, one recovery objective and one difficult dependency to the initial consultation.