On the day of an incident, such as a ransomware attack or a failure of the main storage, everyone is under heavy pressure. Managers ask when work will resume, staff don’t know what to do, and the IT team must decide under pressure which server to restore first, where the passwords are and where replacement hardware will come from. A disaster recovery plan (DR plan) moves these decisions from the day of the crisis to calmer days.
A DR plan is the technical part of a business continuity plan, and it focuses on one question: exactly how, and in what order, do we bring IT systems back?
1. Define scenarios and scope
The plan should cover several realistic scenarios, not just “everything broke.” Examples:
- Failure of an important server or virtual machine
- Failure of the storage or a virtualization host that takes down several servers at once
- A ransomware attack that encrypts the servers
- Complete loss of the office or server room, for example due to fire or an extended power outage
2. Priorities and recovery order
Group systems into priority tiers based on RPO and RTO, and write the recovery order based on dependencies. A typical order is:
- Network, firewall and internet
- Storage and virtualization hosts
- Domain controller and DNS
- Databases and finance software
- File server, email and other services
For each system, note where its backup is, which tool restores it, and how to confirm it works correctly after the restore.
3. Decide on the recovery site in advance
If the primary hardware is destroyed, where will you bring the servers up? Options, from cheapest to most expensive:
- Cold site: You only have backups and procure hardware during the crisis; cheap but slow.
- Warm site: You have hardware or virtualization capacity ready at a branch or data center, and the backups are already there.
- Hot site: Critical servers are continuously replicated and ready to run within a short time.
For many small and mid-sized businesses, a warm site in a data center or second branch offers a good balance between cost and speed.
4. Keep critical information outside your systems
A plan stored only on the file server is not available on ransomware day. Keep a printed copy and a secure offline copy of the following:
- Emergency passwords for admin accounts, the backup console and backup encryption keys
- Contact details for the IT team, managers, contractors, internet provider and data center
- Network diagram, IP addresses and key settings
- License information and support contracts
Define roles: who declares a disaster, who leads the technical recovery, and who communicates with staff and customers. Assign a backup person for each role.
5. Practice and update
A plan that has never been practiced will not work on the day of a crisis. There are three common levels of testing:
- Tabletop exercise: The team walks through a scenario together and reads the steps; gaps in the plan show up quickly.
- Partial restore: One or more important servers are restored in an isolated environment and timed.
- Full exercise: Critical services are brought up at the alternate site, at a time that causes the least disruption.
Update the plan after every major infrastructure change, such as a new server or a migration to another host.
Checklist
- The main scenarios and the plan’s scope are documented.
- The recovery order is written based on dependencies and priorities.
- An alternate recovery site is defined and ready in advance.
- Passwords, keys and contact details are available offline.
- Roles and backup people are assigned.
- The plan is tested and updated at least once a year.
A DR plan doesn’t need to be hundreds of pages; a few precise, tested pages are enough. ITPlus724 engineers can help your company prepare and test this plan, including setting up a recovery site in the data center.