Disaster Recovery Planning (DRP): Beyond the Backup

Let’s maintain the auditor’s mindset for another session. We’ve already seen vulnerabilities, threats, and risks, and even a high-level view of Business Continuity Planning (BCP). Now, we will see the third phase of BCP in depth: Recovery. This is where Disaster Recovery Planning (DRP) takes center stage.

Reference link of BCP

What is Disaster?

To understand recovery, we must first define the catalyst: What is a disaster?

In a professional context, a disaster is any sudden, unplanned event, whether natural, accidental, or malicious, that causes significant damage or loss to an organization’s infrastructure. It is the moment where standard operating procedures are no longer enough to keep the lights on.

And this can lead to a total disruption of the organization. When critical systems fail, the impact ripples through every department. Operations stall, revenue stops, and reputation takes a hit. The DRP is the blueprint designed to navigate this chaos and restore order.

While BCP is the broad umbrella focused on keeping the entire business operational, DRP is the technical subset. It focuses specifically on the IT infrastructure, restoring servers, data, and connectivity after a catastrophic event.

At this stage, the focus shifts from how work continues to how data, systems, and connectivity are restored.

Technical Benchmarks: Metrics

Before looking at backups or servers, recovery success is evaluated using predefined metrics established during the Business Impact Analysis (BIA).

For example; A bank might have an RPO of 0 seconds (no data loss allowed) and an RTO of 2 hours. This necessitates synchronous data mirroring. A local bakery might have an RPO of 24 hours, meaning a simple nightly backup is sufficient.

Recovery Site Strategies: Where do we go?

If the primary data center is a total loss, where does the data go? Recovery site selection is validated against RTO requirements, whether through internal facilities or subscription-based services.

1. Mirror Site (Dual-Active): The most expensive option. Two data centers running simultaneously. If Site A fails, Site B takes the full load instantly with zero downtime.
   

2. Hot Site: A fully equipped facility with a mirrored copy of the data. Recovery happens in minutes or hours.
   

3. Warm Site: Contains the necessary hardware (servers, switches) but no live data. You must restore backups before you can go live. This usually takes 12–24 hours.
   

4. Cold Site: Just a room with power and cooling. You have to ship in hardware and install everything from scratch. This is a “weeks, not days” solution.
   

5. Cloud-Based (DRaaS): Leveraging AWS, Azure, or GCP to spin up virtual instances of your local servers. This is becoming the gold standard for cost-effective, high-speed recovery.
   

Classification of Disaster Levels

Not every technical glitch is a disaster. Auditors look for a classification system that prevents “emergency fatigue.”

• Level 1 & 2 (Minor/Localized): A single server or network switch fails. These are handled by standard IT operations using redundancy (RAID, dual power supplies).

• Level 3 (System Disruption): A critical system, like the ERP or payroll, goes down. This requires specific recovery procedures but not a site move.

• Level 4 (Major Failure): Loss of multiple critical systems or a localized physical issue (e.g., a burst pipe in the server room).

• Level 5 (The True Disaster): The primary site is inaccessible or destroyed. This triggers the full DRP and the migration to an alternate site.

Backup Strategies

Before you can recover, you must have the data. The DRP must specify which backup type is used to meet the Recovery Point Objective (RPO).

Two Way DRP Process

A professional DRP procedure follows a logical flow. An auditor verifies that the documentation covers both the “leaving” and the “coming back” phases:

A) The Failover (Leaving the Main Site):

• Activation: The disaster is declared, and traffic is rerouted to the recovery site.

• Restoration: Data is loaded onto alternate hardware.
  

B) The Failback (Coming Home):

• Reverse Sync: Since new data was created at the recovery site during the disaster, you must sync that “new” data back to the primary site before moving.

• The Switch: Once the primary site is verified as stable, users are moved back. Never shut down your recovery site until the primary site has been running successfully for at least 24–48 hours.

The DRP Execution Procedure (The “Failover”)

This is the process of leaving the main site. It must be documented in a step-by-step manual so that any admin can execute it under pressure.

1. Declaration: The Disaster Recovery Coordinator officially activates the DRP.

2. Evacuation/Safety: Ensure human life is safe before touching the servers.

3. Site Transition: Traffic is rerouted (via DNS or BGP) to the recovery site.

4. Data Restoration: The latest backups (Full + Differential/Incremental) are loaded onto the recovery hardware.

5. Verification: A “Sanity Test” is performed to ensure users can log in and data is consistent.

The Restoration Procedure

Coming back to the main site is often more complex than leaving. This is where most organizations experience data loss.

Phase A: Repair and Prep

• The primary site is repaired (new hardware, cleaned environment).

• The Delta Problem: While you were at the backup site, new data was created. You cannot just “go back” to the old site, or you will lose all work done during the disaster.

Phase B: Reverse Synchronization

• Data is synced from the Recovery Site back to the Primary Site.

• This ensures the primary site now has the most current records.

Phase C: The Switch

• A scheduled maintenance window is set.

• Users are disconnected from the recovery site.

• One final sync is performed.

• Traffic is pointed back to the primary site.

• Post-Mortem: The auditor reviews the logs to see if the RTO and RPO were actually met.

Elements of Recovery Strategy

1. Business Recovery Strategy

Recovery of business operations: It addresses how the organization will fulfill its core mission while the primary systems are down.

• Manual Workarounds: If the digital accounting system is down, can the team process invoices on paper temporarily?

• Criticality Mapping: Identifying which business units (e.g., Customer Support vs. Research) must be brought back first to prevent total collapse.
  

2. Facility & Supply Recovery Strategy

Facility restoration and alternate site enablement. If the primary office or data center is physically inaccessible, you need a place to go. This element focuses on the “bricks and mortar.”

• Alternate Sites: Activating your Hot, Warm, or Cold sites.

• Logistics & Supplies: Ensuring that the alternate site has the necessary physical resources like desks, chairs, specialized hardware, and even basic office supplies to function.

• Lead Times: Managing the time, it takes for vendors to ship replacement hardware to the new location.
  

3. User Recovery Strategy

People and accommodations. This is the most overlooked element. Systems don’t run themselves; people do.

• Crisis Communication: How do you reach employees if the internal email server is down?

• Remote Work/Relocation: Providing the tools (VPNs, laptops) for users to work from home or transporting “key personnel” to the alternate recovery site.

• The Human Element: Addressing employee safety, trauma, and basic needs during a crisis.

4. Technical Recovery Strategy

Recovery of IT services. This is the “engine room” of the DRP. It involves the restoration of the actual technology stack.

• System Restoration: Reinstalling operating systems, configuring network paths (DNS/IP changes), and re-establishing security protocols (Firewalls/MFA).

• Service Order: Bringing up the “Tier 0” services (Active Directory, Identity Management) before trying to launch user-facing applications.

5. Data Recovery Strategy

Recovery of information assets. Data is the lifeblood of the organization. This element ensures that the information being fed into the restored IT services is accurate and current.

• Backup Integrity: Utilizing Full, Differential, or Incremental backups to meet the RPO (Recovery Point Objective).

• Off-site Controls: Ensuring backups were stored far enough away to survive the disaster.

• Data Consistency: Checking that databases are synchronized and haven’t been corrupted during the crash or the restoration process.
  

Testing

A DRP is just a “paper plan” until it is tested. I look for three levels of testing in the audit logs:

1. Tabletop Exercise: A verbal walkthrough of the plan with key stakeholders.

2. Simulation/Parallel Test: Recovering systems at the backup site without stopping the primary site.

3. Full Interruption Test: The gold standard. Actually, shutting down the primary site to prove the backup site can handle the live load.