Cohesity & IBM validate cloud recovery with Terraform

Cohesity said HashiCorp, an IBM company, has validated interoperability between Cohesity Data Cloud's cloud rebuild service and Terraform Self-Managed. It is the first cloud recovery method IBM has cleared for use with Terraform Self-Managed.

The announcement focuses on cyber recovery for cloud applications after destructive attacks. According to Cohesity, Terraform can recreate an organisation's declared cloud configuration from infrastructure-as-code files, while Cohesity Data Cloud restores the latest usable application data from backup.

The companies are positioning the approach as an alternative to manual rebuilds or restoration methods that recreate a production environment as it existed immediately before an attack. In those cases, the restored environment may still include attacker-introduced changes or other errors present at the time of compromise.

Cohesity said its cloud rebuild process instead uses an organisation's reviewed and approved Terraform code as the recovery reference point. That means infrastructure and configuration are rebuilt to the customer's declared state, with data then restored from a selected backup.

The development comes as businesses assess whether existing incident response and recovery plans are suited to attacks that use artificial intelligence. Cohesity cited findings from its 2026 Global Cyber Resilience Report, based on a Vanson Bourne survey of 3,200 security and IT decision-makers, showing that 83% of organisations said their recovery plans would need moderate or significant changes to deal with attacks that use AI to identify and exploit software weaknesses. Only 3% said their plans were ready as they stand.

How it works

Under the process described by Cohesity, cloud rebuild first reads an organisation's Terraform code and maps the cloud resources that need protection, along with the order in which they should be restored. It then backs up both the Terraform code and application data over time, allowing teams to choose a recovery point that links the intended configuration with backed-up data.

When recovery starts, the system recreates the declared infrastructure and configuration from the stored Terraform code and restores application data from the latest usable backup associated with that recovery point. For database services such as Amazon RDS, administrators are guided through a multistep process to reconnect the restored database to the rest of the application, Cohesity said.

The product also shows any timing gap between infrastructure and data backups when they have been captured on different schedules. This gives teams visibility into which data backup will be used with a chosen code version during restoration.

Cohesity added that cloud rebuild can be integrated into existing incident response and recovery workflows. The service also runs customers' Terraform code unchanged through Terraform's official interface.

Vasu Murthy, Chief Product Officer at Cohesity, said the rise of AI-assisted attacks has changed how quickly organisations need to restore systems after a compromise. "Attackers using AI move faster than any team can rebuild by hand. Organisations have already codified exactly what their cloud environments should look like in Terraform. Cohesity uses that code to recreate the declared configuration, then restores the latest usable application data from backup. Teams need both to bring the application back. If you build as code, you can now recover as code," Murthy said.

IBM said customer demand has focused not only on rebuilding infrastructure with Terraform, but also on using Terraform state within a wider recovery process.

"Our customers have asked how they can recover from an attack leveraging Terraform and its state, and not just build a new environment," said Tyler Lynch, Outbound CTO, Automation, IBM Software. "We worked with Cohesity to validate how our technologies work together when they use Terraform state to recover cloud applications, the infrastructure they run on, and the data as part of the recovery process."

The product is available as part of Cohesity Data Cloud Enterprise Edition. Existing users of Data Protect or Cohesity NetBackup can upgrade to that edition to access cloud rebuild, the company said.

Cohesity said the service is designed for organisations that already manage cloud infrastructure through infrastructure-as-code practices and want recovery processes to follow the same model. The validation with Terraform Self-Managed gives those customers a way to restore both declared cloud configuration and backed-up application data through a single recovery method.