Procurement Documents

Disaster Recovery Plan Template for IT (Free, Word)

A free, vendor-neutral DR plan template for IT in Word, with a system inventory, RTO and RPO by tier, a contact tree, step-by-step recovery procedures, failback and a testing log.

20 pages Microsoft Word Updated November 15, 2026
Download Template
Disaster Recovery Plan Template for IT (Free, Word) preview
Disaster Recovery Plan Template for IT (Free, Word)

Preview the first pages

See the structure before you download. The Word file has every section listed below, ready to fill in.

What's inside

The sections in the document, in order.

  1. Plan details and version control
  2. Purpose, scope and assumptions
  3. Roles and contact tree
  4. System inventory and recovery targets
  5. Backup and replication summary
  6. Activation criteria and declaration
  7. Recovery procedures
  8. Communications plan
  9. Providers: contacts and contract terms
  10. Failback

Use this disaster recovery plan template to write down how your IT team brings systems and data back after a serious outage: who can declare a disaster, who does what, which systems come back first, how each one is recovered, and how you prove the plan works. It starts from the same groundwork as the disaster recovery RFP guide: a business impact analysis, your systems listed by recovery tier, and a recovery time objective (RTO) and recovery point objective (RPO) for each tier.


It's written for the IT manager or recovery lead who owns the plan, with sections for system owners, technical leads, communications and the service desk to fill in. If you're also buying recovery services, the plan and the RFP feed each other. The inventory and targets in this plan are what providers price in the DRaaS RFP template, and once the contract is signed, the provider's declaration process, testing terms and contacts go back into the plan.


The Word file is a complete IT disaster recovery plan, with a recovery procedure to copy for each system, failback steps, a testing schedule with a results log, and plan maintenance. It follows the three phases in NIST SP 800-34 Rev. 1, the federal contingency planning guide: activation and notification, recovery, and reconstitution, which this template covers under failback. It also follows ready.gov's IT disaster recovery plan guidance, and it's written for any organization, not only federal agencies.


What's inside this template


  • Plan details, version control and a distribution list, so everyone works from the current copy and can reach it when your network is down
  • Purpose, scope and assumptions, plus roles for the declaring authority, recovery lead, technical leads, security, communications and provider contacts, with a contact tree
  • A system inventory with each system's tier, RTO, RPO, dependencies and recovery method, a recovery order, and a backup and replication summary checked against each RPO
  • Activation criteria, declaration steps with a declaration record, and a recovery log
  • A step-by-step recovery procedure to copy for each system, with escalation rules, and a communications plan by audience and channel
  • Provider contacts and the recovery contract terms the team needs during an outage, from the committed RTO and RPO to test rights and failback charges
  • Failback steps, an after-action review, a testing schedule and results log for tabletop, restore and failover tests, update triggers and a known gaps log

Download the editable Microsoft Word version below. The full text is on this page, so you can read it before you download.

How to use this template

Write one plan for the organization, with a copy of Section 7 for each system. Fill in Sections 1 to 6 first, then the recovery procedures, then test the plan and record the results in Section 11.

  1. Start from the business impact analysis. List the business processes your systems support, and ask each process owner how long the process can stop and how much data it could lose. If you don’t have a business impact analysis yet, NIST publishes a free BIA template with SP 800-34 Rev. 1. The answers set the tiers and targets for everything that follows.
  2. Build the inventory and set the targets. Give each system a tier, an RTO, an RPO and its dependencies, and have its owner agree them. Order the recovery so that shared services such as the network, DNS and identity come back first. Then compare each system’s backups with its RPO in Section 5, and record any gap.
  3. Name the people and the declaration rules. Fill in every role with a primary and an alternate, build the contact tree, and name one declaring authority with successors in a set order. Agree the activation criteria in advance, so the decision to declare is quick and the recovery clock starts at a recorded time.
  4. Write a procedure for each system. Copy Section 7.1 for each system and write the steps so an alternate who has never done the recovery could follow them, from checking dependencies to the system owner confirming it works. Record where the credentials are held, and add the provider contacts and contract terms in Section 9.
  5. Test it, then keep it current. Run tabletop exercises, restore tests and failover tests at the frequency you set for each tier, and log the actual recovery times and data age against the targets. Fix what the tests find, record gaps you can’t close yet, review the plan after every significant change, and replace the offline copies when it changes.

RTO and RPO: how to set them

The recovery time objective (RTO) is the longest a system can stay down after a disruption before the outage does unacceptable harm to the work that depends on it. The recovery point objective (RPO) is how much data you can afford to lose, measured as time back from the disruption to the last usable copy. NIST SP 800-34 Rev. 1 defines both, along with the maximum tolerable downtime (MTD): the total time a business process can be disrupted before the organization suffers significant harm.

RTO decides how much of the recovery environment has to be running already, and RPO decides how often data is backed up or replicated. Tighter targets cost more, so set them from evidence rather than by feel:

  • Start with the business process, not the system. In the business impact analysis, ask each process owner how long the process can stop and how much work could be re-entered or re-created if it were lost.
  • Set each system’s RTO shorter than the MTD of the process it supports. Checking data and re-entering lost work take time after the system is back, and that time counts against the MTD.
  • Set the RPO from how much data the process can lose and how that data would be re-created, for example from source documents or from partners who sent it.
  • Check dependencies. A system can’t come back before the identity, network, DNS and databases it relies on, so they need an RTO at least as short.
  • Group systems into a few tiers, such as critical, important and deferrable, instead of giving every system its own targets. Putting everything in the top tier makes the whole estate cost what your most critical systems cost.
  • Compare the cost of meeting each target with the impact of missing it, and have the business owner agree the result.
  • Prove the targets in a test. Record the actual recovery time and the age of the data recovered. Where a target can’t be met yet, record the gap, the plan to close it and who accepted the risk in the meantime.
  • Treat a provider’s RTO and RPO as a separate question. Check whether each is a contractual service level with remedies or only a target, and when its clock starts. The disaster recovery RFP guide covers how to ask for them.

The full template

This is the complete text of the Word document. Fill in the [bracketed] placeholders and delete the italic guidance before you use or send it.

Instructions: replace every [bracketed placeholder], replace the
example rows with your own, delete rows that don’t apply, and delete the
italic guidance when you’re done. Fill in Sections 1 to 6 first, copy
Section 7 once for each system in Section 4, and record every test in
Section 11.

Base the recovery targets on a business impact analysis (BIA),
which finds out which business processes depend on each system and how
long each process can tolerate an outage. NIST publishes a free BIA
template with SP 800-34 Rev.
1, Contingency Planning Guide for Federal Information Systems
, and
ready.gov explains the method on its business
impact analysis page
.

1. Plan details and version
control

Item Details
Organization [Legal name]
Plan owner (recovery lead) [Name and title]
Declaring authority [Name and title. Successors in Section 6.2]
Approved by [Name, title and date]
Version and effective date [Version], [date]
Next full review due [Date]
Related plans [Business continuity plan, incident response plan, crisis
communications plan: owners and where they’re kept]

1.1 Version history

Version Date Changed by What changed Approved by
[Version] [Date] [Name] [First issue, or the sections changed and why] [Name]

1.2 Distribution

Keep copies the recovery team can reach when your network, email
and file shares are down: printed, offline on an encrypted drive, and at
the recovery site. The plan holds phone numbers and system layouts, so
mark it confidential, control who gets a copy, and swap old copies for
new ones at every update.

Copy holder Format Where it’s kept Version held
[Name, recovery lead] [Print / offline electronic] [Location] [Version]
[Name, declaring authority]
[Recovery site or provider]

2. Purpose, scope and
assumptions

2.1 Purpose

This plan sets out how [Organization] recovers the IT systems and
data listed in Section 4 after a disruption that normal incident
handling can’t fix within a system’s recovery time objective. It names
who can declare a disaster and who does what, and it sets the order of
recovery, the procedure for each system, communications, failback,
testing and upkeep.

2.2 Scope

Item Details
In scope [Sites, data centers, cloud accounts and regions, SaaS data,
networks and end-user services]
Out of scope [What this plan doesn’t cover, and which plan does]
Recovery site or provider [Name and type, e.g. a second data center, a cloud region, a DRaaS
provider, a hot site (fully equipped and running), a warm site (partly
equipped) or a cold site (power and space, but no equipment)]
Disruptions planned for [e.g. loss of the primary site or cloud region, ransomware, loss of
a critical provider, an extended power or network outage]

2.3 Assumptions

List what the plan relies on, so the team knows to adapt when an
assumption proves wrong. Edit these examples.

  • The people in Section 3, or their alternates, can be reached and can
    work from [location, or remotely].
  • The backups in Section 5 are current, readable and stored away from
    the primary site.
  • The recovery site or provider has the capacity its contract commits
    to.
  • At least one way to sign in works if the main identity provider is
    down: [break-glass accounts held in [location]].
  • Staff safety comes first. No one enters an unsafe building to
    recover equipment.

3. Roles and contact tree

3.1 Roles

Name a primary and an alternate for every role. In a small team
one person can hold several roles, but no one should be the only person
who can do a critical step.

Role Responsibilities Primary: name and phone Alternate: name and phone
Declaring authority Decides whether to declare a disaster and when to stand down
Recovery lead Runs the recovery, sets priorities with system owners, owns this
plan between events
Infrastructure and network lead Network, DNS, firewalls, connectivity to the recovery site, compute
and storage
Identity and access lead Directory, single sign-on, break-glass and privileged accounts
Application and database leads Recover and test each application and its data, following Section
7
Security lead Works with incident response when the cause may be a cyber attack,
approves clean recovery points
Communications lead Messages in Section 8, with the approved spokesperson
System owners Agree recovery targets, run manual workarounds, confirm each
recovered system is usable
Provider liaison Declares to the providers in Section 9, tracks their actions and
escalates
Scribe Records times, decisions and actions in the recovery log (Section
6.4)

3.2 Contact tree

Each person calls the people below them, and the recovery lead
confirms that everyone has been reached. If the cause may be a cyber
attack, treat email and chat as possibly down or watched by the
attacker, and use phone calls or the out-of-band channel in Section
8.2.

Who calls Whom Primary contact Backup contact If no answer within [time]
Whoever detects the problem Recovery lead [Phone] [Phone or text] Call the alternate recovery lead
Recovery lead Declaring authority Call the first successor
Recovery lead Technical, security and communications leads Call each alternate
Technical leads Their teams and the system owners
Provider liaison The providers in Section 9 [Number or portal from the contract] Use the provider’s escalation contact

4. System inventory and
recovery targets

Take each system’s tier, RTO and RPO from the BIA, and have the
system owner agree them. The recovery time objective (RTO) is the
longest a system can stay down after a disruption before the outage does
unacceptable harm to the work that depends on it. The recovery point
objective (RPO) is how much data you can afford to lose, measured as
time back from the disruption to the last usable copy. A system can’t
come back before the systems it depends on, so each dependency needs an
RTO at least as short as the systems that rely on it.

4.1 Recovery tiers

Tier What it means RTO RPO
Critical [Processes that stop, or cause serious harm, when the system is down
beyond the RTO, e.g. taking orders, paying staff]
[time] [time]
Important [Processes that can run on manual workarounds for a while] [time] [time]
Deferrable [Systems that can wait until critical and important systems are
back]
[time] [time]

4.2 Systems

Recovery methods include failover to a replica at the recovery
site, restoring from backup, rebuilding from a standard image or
infrastructure-as-code template and then restoring the data, recovery by
a provider under contract, and a manual workaround until the system is
back.

System Owner Tier Supports RTO RPO Depends on Recovery method
[Example: directory and single sign-on] [Name] Critical [Sign-in to every other system] [time] [time] [Network, DNS] [Failover to the replica at [recovery site]]
[Example: finance system] [Name] [Tier] [Invoicing, payments] [time] [time] [Directory, database server] [Restore from backup at [recovery site]]
[Example: SaaS email] [Name] [Tier] [Communication] [The provider’s commitment] [Your own backup’s RPO] [Internet, directory] [The provider restores the service; restore data from your own
backup]

4.3 Recovery order

Shared services such as the network, DNS and identity come back
first. Systems in the same step can be recovered at the same
time.

Order Systems Lead
First [Network, connectivity to the recovery site, DNS] [Infrastructure lead]
Second [Directory, single sign-on, privileged access] [Identity lead]
Third [Core databases and storage] [Database lead]
Fourth [Critical applications] [Application leads]
Then [Important systems, then deferrable systems]

5. Backup and replication
summary

Compare each system’s backups with its RPO, and its tested
restore time with its RTO. Replication alone isn’t a backup, because it
copies corruption and encrypted files to the replica too.

System or data Method and frequency Retention Where copies are kept Offline or immutable copy Last restore test and result Meets RPO
[System] [Backup / replication / snapshot / SaaS backup], every
[interval]
[Period] [Locations away from the primary site] [Yes, where / No] [Date, time taken, age of data restored] [Yes / No: gap in Section 12.2]

5.1 Backup checks

Check Done Owner Date
Critical data has an offline, encrypted backup, and recovery tests
check that it’s available and intact
☐
At least one copy can’t be changed or deleted before its retention
ends, and deleting backups needs more than one person’s approval
☐
Copies are stored far enough from the primary site that the same
event is unlikely to affect both
☐
Backup systems and their administrator accounts are separate from
production credentials, so an attacker who takes over production can’t
delete the backups
☐
Standard images or infrastructure-as-code templates for critical
systems are current and kept with the offline backups, along with
installation media and license keys
☐
The encryption keys and passwords needed to read the backups can be
reached when the primary site and the main password store are down
☐
For each SaaS system, the contract’s restore terms are recorded, and
data the provider won’t restore for you is backed up separately
☐
Where a provider runs your backups, the contract requires it to
follow these practices
☐

6. Activation criteria
and declaration

6.1 Activation criteria

The plan is activated when one or more of these is met. Adjust
them to your systems, and keep them specific enough that the declaring
authority can decide quickly.

  • A critical system is down, or expected to be down, for longer than
    its RTO.
  • The primary site, data center or cloud region is unavailable, unsafe
    or can’t be reached.
  • Data is corrupted or encrypted, for example by ransomware, and can’t
    be fixed within the system in time.
  • A provider service that a critical system depends on is down, and
    the provider can’t give a restore time that fits the RTO.
  • The incident response lead asks for systems to be recovered into a
    clean environment.

6.2 Who can declare

One person holds the authority to declare at any time, with
successors in a set order if that person can’t be reached.

Order Name and title Phone Backup contact
Declaring authority
First successor
Second successor

6.3 Declaration steps

Step Action By Done (time)
Detect Whoever detects the problem tells the recovery lead, using the
contact tree
Service desk, monitoring or a provider
Assess Within [time], assess the cause, the risk of further damage, the
state of the building, power and network, what’s affected, and how long
a normal restore would take
Recovery lead, technical leads
Decide The declaring authority decides to declare a disaster, to recover
named systems only, or to handle it as an incident and reassess at
[time]
Declaring authority
Record Record the time of the declaration. Recovery times are measured from
this point
Scribe
Notify Notify the teams through the contact tree, and declare to providers
by the method in Section 9, recording each provider’s reference
Recovery lead, provider liaison
Check for an attack If a cyber attack is possible, the security lead and incident
response team decide which recovery points are clean and where to
recover, so nothing is restored where the attacker may still have
access
Security lead
Start Start recovery in the order in Section 4.3 Recovery lead

6.4 Declaration record
and recovery log

Item Details
Problem detected [Date, time and by whom]
Disaster declared [Date, time, by whom, and the criteria met]
Scope [Systems and sites]
Provider declaration references [Provider, reference and time]
Stood down [Date, time and by whom]
Time Action or decision By Notes

7. Recovery procedures

Copy Section 7.1 for each system in Section 4.2, in the recovery
order. Write each step so that an alternate called in at night could
follow it, and don’t leave out a step because it seems obvious. Restore
the operating system or platform first, then the application, then its
data.

7.1 Procedure: [System name]

Item Details
Tier, RTO and RPO [Tier], [time], [time]
Recovery method and location [Method], [recovery site or provider]
Technical lead and alternate [Names]
Must be running first [Dependencies from Section 4.2]
Credentials needed and where they’re held [e.g. break-glass account in [location]]
Manual workaround while it’s down [What the system owner’s team does in the meantime]
Last tested [Date and result, from Section 11.3]
Step Action By Expected time Done (time)
1 Confirm the systems this one depends on are running and checked [Name] [time]
2 Choose the recovery point: the most recent copy that’s complete and,
after a cyber attack, confirmed clean by the security lead. Record its
date and time
[Name] [time]
3 Prepare the recovery environment: [compute, storage, network
segment]
[Name] [time]
4 Restore or rebuild the operating system or platform [from image or
template]
[Name] [time]
5 Install or restore the application and its configuration [Name] [time]
6 Restore the data to the chosen recovery point [Name] [time]
7 Reconnect: IP addresses, DNS records, certificates, firewall rules
and integrations [list them]
[Name] [time]
8 Test functions, data and security controls: [e.g. record counts
match, a test transaction posts, sign-in works with multi-factor
authentication, logging is on]
[Name] [time]
9 The system owner confirms the system is usable. Record the time:
this is the actual recovery time to compare with the RTO
[System owner]
10 Tell users and the service desk (Section 8), and end the manual
workaround
[Communications lead]
11 Start backups of the recovered system and confirm the first one
completes
[Name] [time]

Escalation: if a step runs [time] past its expected time, or
fails, the technical lead tells the recovery lead at once. The recovery
lead decides whether to escalate to the provider, bring in more people,
buy or ship equipment, or stay on the workaround, and records the
decision in the log.

8. Communications plan

8.1 Audiences

Audience What they need Who tells them Channel and timing
Recovery teams and leadership Assignments, priorities, expected recovery times, decisions
needed
Recovery lead Bridge call at declaration, then every [interval]
Staff What’s down, workarounds, when the next update comes Communications lead [Text alerts, intranet, phone]
Customers What’s affected and when the next update comes Communications lead, with the spokesperson [Status page, email, phone]
Suppliers and partners whose systems connect to yours What’s affected, and changes to connections or data feeds Provider liaison, technical leads
Regulators and others legal counsel names Notices required by law or contract [Legal counsel] [As required]
Media Approved statements only [Approved spokesperson]

8.2 Channels

Channel Use Details
Bridge call Recovery team coordination [Number and access code, kept offline]
Out-of-band channel When email and chat are down or may be watched by an attacker. Don’t
rely on email alone in any case, because you can’t confirm a message has
been read
[e.g. phone calls, text alerts, or a messaging service separate from
your main accounts]
Status page or hotline Staff and customer updates [Address or number]

Draft the first notice, the update and the all-clear in advance,
and keep them with the plan: [where the templates are kept]. Each
message says what happened, what’s affected, what’s being done, the
expected recovery time if it’s known, what the reader should do, and
when the next update will come. Only the approved spokesperson speaks to
the media. Legal counsel decides which notices laws or contracts
require, such as a data breach notice, and by when. This plan isn’t
legal advice.

9. Providers: contacts
and contract terms

List every provider the recovery depends on. If you’re choosing a
recovery provider, RFPhub’s free DRaaS
RFP template
and the disaster
recovery RFP guide
cover what to require. Once the contract is
signed, copy its terms below so no one has to search for it during an
outage. Where this summary and the contract differ, the contract
governs, and legal counsel reviews contract terms.

9.1 Provider contacts

Provider Service How to declare or open a priority case Contract or account ID Escalation contact Who on your side may declare
[DRaaS or recovery site provider] [Service model: self-service, assisted or managed] [Phone answered at any hour, or portal]
[Backup provider]
[Cloud or hosting provider]
[Internet and network carriers]
[SaaS providers for critical systems]
[Hardware supplier, offsite media storage, managed service
provider]

9.2 Recovery provider
contract terms

Term What the contract says Contract section
The RTO and RPO committed for each tier, and whether each is a
contractual service level with remedies or only a target
When the recovery clock starts and stops
How to declare, who may declare, and any declaration fee
Recovery capacity reserved for you, and what happens if many
customers declare at once
How long you can run at the recovery site, and the rate charged once
you declare
Tests included, the notice to give, and the price of each extra
test
Failback support, and charges for moving data back
Security incident notification: the time allowed and the
contact
Data return and deletion at the end of the contract
Service credits: the amount, how to claim and the deadline

10. Failback

Failback moves systems from the recovery site back to the primary
site, or to a new permanent home, with the changes made during the
outage, once that site is fixed. Plan it as a change at a time the
system owners agree, because a failed failback is a second
outage.

10.1 Failback steps

Step Action By Done (time)
1 Confirm the primary site is ready: repaired and tested, the cause
fixed and, after a cyber attack, confirmed clean by the security
lead
Infrastructure lead, security lead
2 Agree a cutover window with the system owners, tell users, book the
provider’s failback support, and set a rollback plan: stay on the
recovery site and retry at [time]
Recovery lead
3 Back up the systems at the recovery site
4 Copy the data changed during the outage back to the primary
site
5 Check the data is complete and current, e.g. record counts and the
latest transactions
6 Test functions, integrations and security controls at the primary
site
7 Switch users, DNS records and integrations back. Where it’s
practical, run both sites side by side until you’re sure the primary is
working correctly
8 The system owners confirm each system is usable
9 Take a full backup of each restored system
10 Release the recovery environment as the contract allows, so charges
for the declared disaster stop, and return any backup media to offsite
storage
Provider liaison
11 The declaring authority stands down, and everyone in Section 8 is
told
Declaring authority

10.2 After-action review

Hold the review within [time] of standing down, with everyone who
took part. Compare the actual recovery times and data loss with the
targets, record what worked and what didn’t, and update the
plan.

Finding Action Owner Due date Status

11. Testing schedule and
results log

Set the frequency from each system’s tier, test critical systems
more often, and test again after significant changes. Some regulations,
frameworks and contracts set a minimum for the organizations they cover,
so check which apply to you. Agree success criteria in writing before
each test, and test as close to real conditions as you safely
can.

11.1 Test types

Test What it involves What it proves
Contact tree test Run the contact tree, out of hours if you can People can be reached within [time]
Tabletop exercise A facilitator walks the team through a scenario, such as ransomware
or the loss of the primary site. Discussion only: no systems are
touched
Roles and decisions are clear, and gaps show up
Restore test Restore a system or data set from backup into a test environment,
then check it
Backups are readable and complete, and restores fit the RTO and
RPO
Failover test Bring systems up at the recovery site with their dependencies, as in
a real declaration, then fail back
The recovery site, procedures and failback work end to end

CISA publishes free tabletop
exercise packages
with ready-made scenarios, including ransomware
and natural disasters.

11.2 Schedule

Test Systems or tier Frequency Next date Owner
Contact tree test All roles Every [interval]
Tabletop exercise [Scenario] Every [interval]
Restore test Critical Every [interval]
Restore test Important and deferrable Every [interval]
Failover test [Critical systems and their dependencies] Every [interval], and after major changes

11.3 Results log

Time each test from the declaration, or the simulated one, to the
system owner confirming the system is usable, and record the age of the
data recovered.

Date and test Systems Success criteria RTO: target and actual RPO: target and actual Result Actions, owners and dates
[Met / Partly met / Not met]

12. Plan maintenance

Review the whole plan every [interval] and whenever something
significant changes, and check contact details more often than that.
Record every change in Section 1.1, and replace the copies listed in
Section 1.2.

12.1 Update triggers

Trigger What to update Owner
A system is added, replaced or retired Sections 4, 5 and 7, and the BIA System owner, recovery lead
Architecture, hosting, cloud region or network changes Sections 4, 5, 7 and 10 Technical leads
Someone joins, leaves or changes role Sections 1.2, 3 and 6.2 Recovery lead
A provider contract is signed, renewed or changed Section 9 Provider liaison
A test or a real event finds a problem The sections affected Recovery lead
Business priorities change, or the BIA is updated Tiers and targets in Section 4 System owners

12.2 Known gaps

Where a test shows a target can’t be met yet, record it here with
a plan to close the gap, and have the system owner accept the risk in
the meantime.

System Target What the test showed Plan to close the gap Owner Due date Risk accepted by

13.
Disaster recovery plan vs business continuity plan

A business continuity plan keeps business processes going during and
after a disruption, such as paying staff or serving customers, and
covers people, workplaces and manual workarounds. This disaster recovery
plan restores the IT systems and data those processes depend on.
Ready.gov advises developing the IT disaster recovery plan together with
the business continuity plan, with IT priorities and recovery time
objectives set in the business impact analysis. Keep the two consistent:
a system’s RTO should be short enough that the business process
depending on it is back within the time it can tolerate, with room to
check data and re-enter lost work. An incident response plan, which
deals with cyber attacks, can trigger this plan.

Terms vary. NIST SP 800-34 Rev. 1 uses “disaster recovery plan”
for moving systems to an alternate site after a major disruption, and
“information system contingency plan” for recovering a single system
wherever it runs. This template covers both: Sections 6 and 10 run the
whole recovery, and each procedure in Section 7 recovers one
system.

Frequently asked questions

What should a disaster recovery plan include?

It should name who can declare a disaster and who does what, with a contact tree and alternates for every role. It needs an inventory of systems with each one’s tier, RTO, RPO, dependencies and recovery method, a summary of backups, and a step-by-step recovery procedure for each system. It also covers communications, provider contacts and contract terms, failback, a testing schedule with results, and how the plan is kept current.

What is the difference between a disaster recovery plan and a business continuity plan?

A business continuity plan keeps business processes running during and after a disruption, including people, workplaces and manual workarounds. A disaster recovery plan restores the IT systems and data those processes depend on. Both draw on the same business impact analysis, which sets the priorities and recovery times, so build and test them together.

How often should a disaster recovery plan be tested?

Set the frequency yourself from each system’s tier, and test critical systems more often than the rest, using a mix of tabletop exercises, restore tests and failover tests. Test again after significant changes, such as a new critical system, a move to a new hosting provider or a new recovery provider. Some regulations, frameworks and contracts set a minimum for the organizations they cover, so check which apply to you.

Who owns the disaster recovery plan?

Give the plan one owner, such as an IT manager or a named recovery lead, who keeps it current, runs the tests and leads a recovery. One senior manager, such as the CIO, holds the authority to declare a disaster, with named successors. System owners agree the recovery targets for their systems, and technical leads write and maintain each system’s recovery procedure.

What is the difference between RTO and RPO?

RTO, the recovery time objective, is the longest a system can stay down after a disruption. RPO, the recovery point objective, is how much data you can afford to lose, measured as time back from the disruption to the last usable copy. RTO drives how much of the recovery environment has to be ready in advance, while RPO drives how often you back up or replicate data.

Related guides and templates

Download Template

More procurement documents templates