Create a Right-Sized Disaster Recovery Plan

Close the gap between your DR capabilities and service continuity requirements.

ANALYST PERSPECTIVE

An effective disaster recovery plan (DRP) is not just an insurance policy.

"An effective DRP addresses common outages such as hardware and software failures, as well as regional events, to provide day-to-day service continuity. It’s not just insurance you might never cash in. Customers are also demanding evidence of an effective DRP, so organizations without a DRP risk business impact not only from extended outages but also from lost sales. If you are fortunate enough to have executive buy-in, whether it’s due to customer pressure or concern over potential downtime, you still have the challenge of limited time to dedicate to disaster recovery (DR) planning. Organizations need a practical but structured approach that enables IT leaders to create a DRP without it becoming their full-time job."

Frank Trovato,

Research Director, Infrastructure

Info-Tech Research Group

Is this research for you?

This Research Is Designed For:

• Senior IT management responsible for executing DR.
• Organizations seeking to formalize, optimize, or validate an existing DRP.
• Business continuity management (BCM) professionals leading DRP development.

This Research Will Help You:

• Create a DRP that is aligned with business requirements.
• Prioritize technology enhancements based on DR requirements and risk-impact analysis.
• Identify and address process and technology gaps that impact DR capabilities and day-to-day service continuity.

This Research Will Also Assist:

• Executives who want to understand the time and resource commitment required for DRP.
• Members of BCM and crisis management teams who need to understand the key elements of an IT DRP.

This Research Will Help Them:

• Scope the time and effort required to develop a DRP.
• Align business continuity, DR, and crisis management plans.

Executive summary

Situation

• Any time a natural disaster or major IT outage occurs, it increases executive awareness and internal pressure to create a DRP.
• Industry standards and government regulations are driving external pressure to develop business continuity and IT DR plans.
• Customers are asking suppliers and partners to provide evidence that they have a workable DRP before agreeing to do business.

Complication

• Traditional DRP templates are onerous and result in a lengthy, dense plan that might satisfy auditors, but will not be effective in a crisis.
• The myth that a DRP is only for major disasters leaves organizations vulnerable to more common incidents.
• The growing use of outsourced infrastructure services has increased reliance on vendors to meet recovery timeline objectives.

Resolution

• Create an effective DRP by following a structured process to discover current capabilities and define business requirements for continuity:
  • Define appropriate objectives for service downtime and data loss based on business impact.
  • Document an incident response plan that captures all of the steps from event detection to data center recovery.
  • Create a DR roadmap to close gaps between current DR capabilities and recovery objectives.

Info-Tech Insight

1. At its core, DR is about ensuring service continuity. Create a plan that can be leveraged for both isolated and catastrophic events.
2. Remember Murphy’s Law. Failure happens. Focus on improving overall resiliency and recovery, rather than basing DR on risk probability analysis.
3. Cost-effective DR and service continuity starts with identifying what is truly mission critical so you can focus resources accordingly. Not all services require fast failover.

An effective DRP is critical to reducing the cost of downtime

If you don’t have an effective DRP when failure occurs, expect to face extended downtime and exponentially rising costs due to confusion and lack of documented processes.

Potential Lost Revenue

The impact of downtime tends to increase exponentially as systems remain unavailable (graph at left). A current, tested DRP will significantly improve your ability to execute systems recovery, minimizing downtime and business impact. Without a DRP, IT is gambling on its ability to define and implement a recovery strategy during a time of crisis. At the very least, this means extended downtime – potentially weeks or months – and substantial business impact.

Adapted from: Philip Jan Rothstein, 2007

Cost of Downtime for the Fortune 1000

Cost of unplanned apps downtime per year: $1.25B to $2.5B.

Cost of critical apps failure per hour: $500,000 to $1M.

Cost of infrastructure failure per hour: $100,000.

35% reported to have recovered within 12 hours.

17% of infrastructure failures took more than 24 hours to recover.

13% of application failures took more than 24 hours to recover.

Source: Stephen Elliot, 2015

Info-Tech Insight

The cost of downtime is rising across the board, and not just for organizations that traditionally depend on IT (e.g. e-commerce). Downtime cost increase since 2010:

Hospitality: 129% increase

Transportation: 108% increase

Media organizations: 104% increase

An effective DRP also sets clear recovery objectives that align with system criticality to optimize spend

Take a practical approach that creates a more concise and actionable DRP

DR planning is not your full-time job, so it can’t be a resource- and time-intensive process.

DR must be integrated with day-to-day incident management to ensure service continuity

When a tornado takes out your data center, it’s an obvious DR scenario and the escalation towards declaring a disaster is straightforward.

The challenge is to be just as decisive in less-obvious (and more common) DR scenarios such as a critical system hardware/software failure, and knowing when to move from incident management to DR. Don’t get stuck troubleshooting for days when you could have failed over in hours.

Bridge the gap with clearly-defined escalation rules and criteria for when to treat an incident as a disaster.

Source: Info-Tech Research Group; N=92

Myth busted: The DRP is separate from day-to-day ops and incident management.

The most common threats to service continuity are hardware and software failures, network outages, and power outages

Source: Info-Tech Research Group; N=87

Info-Tech Insight

Does this mean I don’t need to worry about natural disasters? No. It means DR planning needs to focus on overall service continuity, not just major disasters. If you ignore the more common but less dramatic causes of service interruptions, you are diminishing the business value of a DRP.

Myth busted: DRPs are just for destructive events – fires, floods, and natural disasters.

DR isn’t about identifying risks; it’s about ensuring service continuity

The traditional approach to DR starts with an in-depth exercise to identify risks to IT service continuity and the probability that those risks will occur.

Here’s why starting with a risk register is ineffective:

• Odds are, you won’t think of every incident that might occur. If you think of twenty risks, it’ll be the twenty-first that gets you. If you try to guard against that twenty-first risk, you can quickly get into cartoonish scenarios and much more costly solutions.
• The ability to failover to another site mitigates the risk of most (if not all) incidents (fire, flood, hardware failure, tornado, etc.). A risk and probability analysis doesn’t change the need for a plan that includes a failover procedure.

Where risk is incorporated in this methodology:

• Use known risks to further refine your strategy (e.g. if you are prone to hurricanes, plan for greater geographic separation between sites; ensure you have backups, in addition to replication, to mitigate the risk of ransomware).
• Identify risks to your ability to execute DR (e.g. lack of cross-training, backups that are not tested) and take steps to mitigate those risks.

Myth busted: A risk register is the critical first step to creating an effective DR plan.

You can’t outsource accountability and you can’t assume your vendor’s DR capabilities meet your needs

Outsourcing infrastructure services – to a cloud provider, co-location provider, or managed service provider (MSP) – can improve your DR and service continuity capabilities. For example, a large public cloud provider will generally have:

• Redundant telecoms service providers, network infrastructure, power feeds, and standby power.
• Round-the-clock infrastructure and security monitoring.
• Multiple data centers in a given region, and options to replicate data and services across regions.

Still, failure is inevitable – it’s been demonstrated multiple times1 through high-profile outages. When you surrender direct control of the systems themselves, it’s your responsibility to ensure the vendor can meet your DR requirements, including:

• A DR site and acceptable recovery times for systems at that site.
• An acceptable replication/backup schedule.

Sources: Kyle York, 2016; Shaun Nichols, 2017; Stephen Burke, 2017

Myth busted: I outsource infrastructure services so I don’t have to worry about DR. That’s my vendor’s responsibility.

Choose flowcharts over process guides, checklists over procedures, and diagrams over descriptions

IT DR is not an airplane disaster movie. You aren’t going to ask a business user to execute a system recovery, just like you wouldn’t really want a passenger with no flying experience to land a plane.

In reality, you write a DR plan for knowledgeable technical staff, which allows you to summarize key details your staff already know. Concise, visual documentation is:

• Quicker to create.
• Easier to use.
• Simpler to maintain.

"Without question, 300-page DRPs are not effective. I mean, auditors love them because of the detail, but give me a 10-page DRP with contact lists, process flows, diagrams, and recovery checklists that are easy to follow."

– Bernard Jones, MBCI, CBCP, CORP, Manager Disaster Recovery/BCP, ActiveHealth Management

Source: Info-Tech Research Group; N=95

*DR Success is based on stated ability to meet recovery time objectives (RTOs) and recovery point objectives (RPOs), and reported confidence in ability to consistently meet targets.

Myth busted: A DRP must include every detail so anyone can execute recovery.

A DRP is part of an overall business continuity plan

A DRP is the set of procedures and supporting documentation that enables an organization to restore its core IT services (i.e. applications and infrastructure) as part of an overall business continuity plan (BCP), as described below. Use the templates, tools, and activities in this blueprint to create your DRP.

Note: For DRP, we focus on business-facing IT services (as opposed to the underlying infrastructure), and then identify required infrastructure as dependencies (e.g. servers, databases, network).

Take a practical but structured approach to creating a concise and effective DRP

Info-Tech offers various levels of support to best suit your needs

DIY Toolkit

"Our team has already made this critical project a priority, and we have the time and capability, but some guidance along the way would be helpful."

Guided Implementation

“Our team knows that we need to fix a process, but we need assistance to determine where to focus. Some check-ins along the way would help keep us on track.”

Workshop

“We need to hit the ground running and get this project kicked off immediately. Our team has the ability to take this over once we get a framework and strategy in place.”

Consulting

“Our team does not have the time or the knowledge to take this project on. We need assistance through the entirety of this project.”

Diagnostics and consistent frameworks used throughout all four options

Info-Tech advisory services deliver measurable value

Info-Tech members save an average of $22,983 and 22 days by working with an Info-Tech analyst on DRP (based on client response data from Info-Tech Research Group’s Measured Value Survey, following analyst advisory on this blueprint).

Why do members report value from analyst engagement?

1. Expert advice on your specific situation to overcome obstacles and speed bumps.
2. Structured project and guidance to stay on track.
3. Project deliverables review to ensure the process is applied properly.

Guided implementation overview

Your trusted advisor is just a call away.

Define DRP scope (Call 1)

Scope requirements, objectives, and your specific challenges. Identify applications/ systems to focus on first.

Define current status and system dependencies (Calls 2-3)

Assess current DRP maturity. Identify system dependencies.

Conduct a BIA (Calls 4-6)

Create an impact scoring scale and conduct a BIA. Identify RTO and RPO for each system.

Recovery workflow (Calls 7-8)

Create a recovery workflow based on tabletop planning. Identify gaps in recovery capabilities.

Projects and action items (Calls 9-10)

Identify and prioritize improvements. Summarize results and plan next steps.

Your guided implementations will pair you with an advisor from our analyst team for the duration of your DRP project.

End-user complaints distract from serious IT-based risks to business continuity

Case Study

Industry: Manufacturing
Source: Info-Tech Research Group Client Engagement

A global manufacturer with annual sales over $1B worked with Info-Tech to improve DR capabilities.

DRP BIA

Conversations with the IT team and business units identified the following impact of downtime over 24 hours:

• Email: Direct Cost: $100k; Goodwill Impact Score: 8.5/16
• ERP: Direct Cost: $1.35mm; Goodwill Impact Score: 12.5/16

Tabletop Testing and Recovery Capabilities

Reviewing the organization’s current systems recovery workflow identified the following capabilities:

• Email: RTO: minutes, RPO: minutes
• ERP: RTO: 14 hours, RPO: 24 hours

Findings

Because of end-user complaints, IT had invested heavily in email resiliency though email downtime had a relatively minimal impact on the business. After working through the methodology, it was clear that the business needed to provide additional support for critical systems.

Insights at each step:

Identify DR Maturity and System Dependencies

Conduct a BIA

Outline Incident Response and Recovery Workflow With Tabletop Exercises

Mitigate Gaps and Risks

Create a Right-Sized Disaster Recovery Plan

Phase 1

Define DRP Scope, Current Status, and Dependencies

Step 1.1: Set Scope, Kick-Off the DRP Project, and Create a Charter

This step will walk you through the following activities:

• Establish a team for DR planning.
• Retrieve and review existing, relevant documentation.
• Create a project charter.

This step involves the following participants:

• DRP Coordinator
• DRP Team (Key IT SMEs)
• IT Managers

Results and Insights

• Set scope for the first iteration of the DRP methodology.
• Don’t try to complete your DR and BCPs all at once.
• Don’t bite off too much at once.

Kick-off your DRP project

You’re ready to start your DR project.

This could be an annual review – but more likely, this is the first time you’ve reviewed the DR plan in years.* Maybe a failed audit might have provided a mandate for DR planning, or a real disaster might have highlighted gaps in DR capabilities. First, set appropriate expectations for what the project is and isn’t, in terms of scope, outputs, and resource commitments. Very few organizations can afford to hire a full-time DR planner, so it’s likely this won’t be your full-time job. Set objectives and timelines accordingly.

Gather a team

• Often, DR efforts are led by the infrastructure and operations leader. This person can act as the DRP coordinator or may delegate this role.
• Key infrastructure subject-matter experts (SMEs) are usually part of the team and involved through the project.

Find and review existing documentation

• An existing DRP may have information you can re-purpose rather than re-create.
• High-level architecture diagrams and network diagrams can help set scope (and will become part of your DR kit).
• Current business-centric continuity of operations plans (COOPs) or BCPs are important to understand.

Set specific, realistic objectives

• Create a project charter (see next slide) to record objectives, timelines, and assumptions.

*Only 20% of respondents to an Info-Tech Research Group survey (N=165) had a complete DRP; only 38% of respondents with a complete or mostly complete DRP felt it would be effective in a crisis.