Organisations can safely restore systems after ransomware by containing the incident, preserving evidence, confirming a trusted recovery source and removing the attacker's access before reconnecting anything. Rebuild or recover systems in a controlled order, test them while isolated, then return services in stages with business owners involved. A backup is one recovery option; it is not proof that restoration is safe.
In This Article
What needs to happen before recovery begins?

“John brought the right knowledge to my issue and resolved it in about a reasonable amount of time. I walked away confident the problem was actually fixed.”
First, contain the incident. Isolate affected systems in a way that limits further spread, but coordinate with incident responders and system owners before shutting down equipment. A system may support clinical work, building access or production, and an abrupt shutdown can create operational or safety risks.
Preserve relevant logs, ransom notes, security alerts and a record of actions taken. Investigators may need those details to establish how the incident began, which accounts and systems were affected, and whether access remains. Do not wipe or rebuild a compromised device until the response team has decided what evidence to retain.
Name an incident lead and agree who can authorize restoration changes. Identify the people responsible for security decisions and for communicating service status. If external responders, insurers, legal counsel or regulators are involved, follow the organisation's established process rather than guessing at notification requirements.
Keep recovery work separate from systems that may still be compromised. Use known-clean administrative workstations and accounts where available. If identity services, privileged accounts or a domain may have been affected, do not rely on them as trusted control points until they have been assessed and secured.
Use four recovery gates: containment, a trusted recovery source, a clean rebuild and verified operation.
Begin restoration only when responders have a reasonable basis to believe the threat is contained. If the attacker's access remains unclear, reconnecting systems can expose the recovery environment to the same incident. Containment comes first.
How can we decide whether a backup is safe to restore?
Confirm that the backup is available, sufficiently complete for the intended recovery and isolated from the compromised environment. Check whether attackers could have accessed backup consoles, administration accounts or storage during the incident. A copy exposed alongside production systems may have been encrypted, altered or otherwise compromised.
Choose a recovery point based on both integrity and business impact. Establish what data and services it contains, when it was created and what work would be lost by using it. The newest copy is not automatically the safest choice if it includes attacker changes, compromised files or an unresolved route back into the environment.
Test the restore in an isolated environment before reconnecting the system. Confirm that the operating system and applications start, required data can be read, and the necessary dependencies are present. A successful backup job only confirms that a job ran; it does not prove that the data can support a working recovery.
If an isolated test environment is not available, pause and have qualified responders assess the risk before proceeding. Record which recovery point was selected and why. Note gaps such as missing transactions, unavailable configurations or data that staff will need to recreate.
Business owners need those details to weigh service priorities against potential data loss. Organisations can review data backup and recovery services alongside disaster recovery planning to clarify how recovery sources, system dependencies and restoration decisions fit together. Plans should be tested against real business workflows, not just infrastructure diagrams.
A backup is a viable recovery source only after its integrity and practical recoverability have been assessed. Treat validation as a required control.
Which systems should come back first?
Set the restoration order according to business dependencies and the consequences of an outage. A common starting point is the services needed to administer and monitor the environment, followed by identity and core infrastructure, priority applications and user endpoints. The right sequence depends on the organisation's own architecture and work processes.
Map what each service needs and what relies on it. A business application may depend on identity, DNS, storage, network access and a database. A building system may rely on local controllers, management software and specific administrator accounts. Bringing back the application without its dependencies can leave staff unable to use it.
Use this sequence as a planning aid, then adapt it to the organisation's dependencies and incident findings:
Restore the clean administrative and monitoring capabilities needed to manage recovery.
Bring back identity and core infrastructure, validating their security and dependencies before proceeding.
Restore priority applications and the data they need, then test representative business workflows.
Reconnect user endpoints and expand access in stages while monitoring for unusual activity.
Include operational leaders in priority decisions. IT can explain technical dependencies; business owners know which workflows can pause, which have safe manual workarounds and what order will restore useful service. Healthcare, legal, financial and property teams should assess the consequences for the work they are accountable for.
Restore a limited group of systems first, test security and function, then expand in controlled stages. Keep monitoring active and agree what would trigger a pause or isolation. Do not reconnect every recovered device at once just because a server is available.
Connected buildings need coordination across IT and building operations. Access control, cameras, tenant connectivity and other services may have separate owners and dependencies. The Built, Wired & Secured article on ransomware containment in connected buildings discusses how an incident can extend into those operational systems.
Choose the sequence that restores safe, useful business capability. Dependency mapping helps teams see which systems must be ready first.
How should teams rebuild and secure systems before reconnecting them?
When compromise is confirmed or cannot be ruled out, prefer a clean rebuild from trusted installation media or a known-good image over returning an affected system to service unchanged. Apply required updates and configuration, then restore only the data and applications needed for the system's approved role. Responders should decide whether a system can be cleaned safely or needs replacement.
Disable accounts, remote-access paths and credentials believed to be exposed. Reset passwords and secrets through a controlled process, starting with privileged and service accounts. Review access rights and confirm that each account has an owner and a defined purpose. Never use a compromised administrator credential to manage a rebuilt environment.
Before reconnecting, check endpoint protection, logging, network controls and monitoring. Confirm that alerts reach a responsible person and that the team can see relevant activity. These checks reduce the chance that a recovered service reintroduces the same access path or operates without oversight.
Keep recovered systems segmented while testing. Limit connections to what each service needs, and ask the system owner to confirm that essential workflows work. If a system communicates with cloud services, other networks or building controls, verify each connection deliberately instead of opening broad access by default.
GDS Technology works across connected technology systems, linking security decisions to business operations and risk. Its ransomware removal service may be relevant when an organisation is assessing recovery support. The response scope and restoration decisions should be confirmed for the specific incident.
A system is not ready to return just because it boots. Security checks and operational tests both need to pass.
How do we know recovery is complete?
Test the full business workflow, not just a server status or login screen. Ask the people who do the work to complete representative tasks, access the right records and use connected services. For example, a property team should test the administrative functions it needs to manage a building; an office team should test the applications and shared resources required for daily work.
Check that restored data matches the selected recovery point and that integrations, scheduled processes and permissions behave as expected. If transactions or records are missing, document the gap and agree on a safe handling process with the business owner. Do not report a complete recovery while known data gaps remain undisclosed.
Monitor recovered services for unusual activity and performance issues. Keep responders involved until the agreed criteria for closing their work have been met. Record what was restored, the source used, the tests performed, open issues and the person who accepted each service back into operation.
A professional services customer described how knowledgeable support and follow-up helped them feel that an issue had been resolved. Madhav, in professional services, said, "I walked away confident the problem was actually fixed." During recovery, that confidence should come from people testing the system and confirming that it supports their work, not simply hearing that it has been switched back on.
After services return, review the incident and update the recovery plan. Record actual dependencies, decision delays, data gaps and workarounds. Assign owners and due dates to corrective actions. If a workaround depends on one person remembering an undocumented step, it is not yet a repeatable recovery process.
Recovery is complete when business owners accept tested services, monitoring remains in place, and unresolved risks and data gaps have named owners. An IT compliance review may help assess documentation and control obligations, but restoration alone does not establish compliance.
A few questions to ask about your own setup:
- Can we verify backups before restoring?
- Do we know which systems are safe to bring back?
- Who decides when restored systems can reconnect?
If you're weighing how your recovery process would work, you can get in touch with our team to talk through your questions.
Frequently Asked Questions
Should we pay a ransom to restore encrypted files?
Do not treat payment as a dependable recovery method. It does not establish that a working decryptor will be provided, that all data can be recovered, or that attackers have lost access. Preserve evidence and involve incident responders, legal counsel and relevant insurers or authorities through established channels. Compare recovery options using verified backups and business impact.
Can we reconnect a restored server as soon as it starts?
No. A server starting successfully does not confirm that it is clean, correctly configured or safe to reconnect. Validate the recovery source, rebuild or remediate as directed by responders, reset exposed credentials, check security controls and test dependencies. Reconnect in a controlled stage after the technical and business owners agree that required checks have passed.
What if the organisation has no tested backup?
Pause before attempting broad restoration. Identify available copies, snapshots and unaffected data, then have qualified responders assess their integrity and the incident's scope. Prioritize services with business owners, document data gaps and consider safe manual workarounds. Avoid overwriting potential evidence or reconnecting systems until the recovery path and risks are understood.
How should we decide which services to restore first?
Use business impact and technical dependencies to set the order. Identify the services required to administer and secure recovery, then map what priority applications and workflows depend on. Ask operational leaders to weigh urgency, safe workarounds and consequences of delay. Restore a limited group first, test the full workflow and expand in controlled stages.
Follow GDS Technology on Facebook, and subscribe on YouTube for new videos.