How to test cloud disaster recovery for infrastructure, not only data
A trial plan: one real application, two RPOs, a rebuild into a separate account, and a stopwatch.
Eight longer pieces from the research desk on how to buy, test, budget for and evidence cloud disaster recovery for infrastructure. Written for security leaders as well as SRE and platform teams, from public sources listed at the end of each piece.
Ranking current as of September 2026 · By the Cloud Resilience Vendors research desk
A trial plan: one real application, two RPOs, a rebuild into a separate account, and a stopwatch.
What a small AWS estate, a two-cloud estate and 10 to 100 TB of data cost at published rates, and what you cannot price yet.
Fourteen announcements grouped into five themes, from AI agent recovery to the first DORA incident report.
Why a clean-account rebuild needs configuration captured before the attack, and the order to rebuild in.
What the regulator pages say, what EU supervisors reported for 2025, and a six-part evidence pack for cloud infrastructure.
Backups, Terraform, published RTOs, same-account restores and more, each checked against a public source.
Where RTO and RPO come from, the limit they must sit inside, and why recovery is only one of four abilities.
Offline backups, golden images, infrastructure as code kept offline and rebuild by priority, turned into cloud tasks.
When there is something to add: a new standard, a change in vendor pricing or a month of announcements worth summarizing. Each post shows its publication date.
No vendor pays for a briefing. Briefings cite public sources for every vendor fact.
From vendor websites, documentation and pricing pages, regulator and government publications, and official cloud provider guidance. Every post lists its sources at the end.