Website Backup and Disaster Recovery Plan: RPO, RTO, and Restore-Test Template
website backupsdisaster recoveryRPORTOWordPress securityhosting security

Website Backup and Disaster Recovery Plan: RPO, RTO, and Restore-Test Template

SSmart Hosting Editorial Team
2026-11-14
7 min read

Build a practical website backup and disaster recovery plan with RPO, RTO, retention, off-site storage, access controls, and restore tests.

A backup is only useful if it contains the right data and can be restored within your business’s recovery limits. This website backup strategy gives you a practical way to define RPO and RTO, choose retention and off-site storage, protect backup access, and test a restore before a server failure, malware incident, or failed update makes recovery urgent.

Overview

A website disaster recovery plan describes how you will bring a site and its supporting services back after data loss or an outage. It should cover more than a copy of the public web files. Depending on the site, recovery may require databases, uploaded media, email, DNS records, SSL configuration, scheduled jobs, environment variables, application settings, and administrator access.

Two targets make the plan measurable:

  • Recovery point objective (RPO): the maximum amount of recent data you can afford to lose. An RPO of 24 hours may be acceptable for a low-change brochure site; a store, membership site, or active publication may need backups several times a day.
  • Recovery time objective (RTO): the maximum acceptable time to restore service. A site that supports internal operations may tolerate a longer recovery window than a site that processes customer orders.

Set these targets before selecting a backup schedule. A backup that runs often but cannot be downloaded, accessed during a hosting-account lockout, or restored to a usable environment does not meet the objective in practice.

Start by inventorying what the website depends on. Record the CMS and version, hosting type, database engine, PHP or runtime version, DNS provider, domain registrar, email provider, CDN, payment integrations, and any external services. For WordPress, include themes, plugins, uploads, the database, and configuration files. A hosting control panel backup may not include every external dependency, so document what must be recovered separately.

Checklist by scenario

1. Choose what to back up

DataWhy it mattersBackup guidanceRestore check
Website filesContains application code, themes, plugins, templates, and static assets.Capture the complete document root and any application directories outside it.Confirm file permissions, ownership, and required runtime versions.
DatabaseUsually contains pages, settings, users, orders, comments, and other changing records.Use a consistent database export and schedule it according to the site’s RPO.Import it into an isolated database and verify key records and application connections.
Media and uploadsMay include product images, documents, user uploads, and editorial assets.Back up separately when its size or change rate makes a full-site copy inefficient.Open representative files and check paths, permissions, and application references.
Email dataMay include business correspondence, mailboxes, contacts, and delivery settings.Confirm whether email is hosted with the web host or a separate provider; use that provider’s export or retention process.Verify mailbox access, aliases, forwarding, and the records needed for delivery.
Configuration and access dataIncludes DNS records, cron jobs, environment variables, API settings, SSL details, and recovery contacts.Store a readable configuration inventory without placing passwords in an unprotected document.Check that credentials, keys, and certificates can be rotated or reissued safely.

2. Match backup frequency to change rate

  • Low-change informational site: use a regular full backup and additional copies before content or platform changes.
  • Frequently updated site: back up the database more often than static files if content changes throughout the day.
  • Commerce or membership site: define how orders, accounts, payments, and transactions will be reconciled if a restore loses recent activity.
  • Development or staging environment: back up only data that needs recovery, but document whether it can be rebuilt from version control and configuration files.

Do not treat frequency as the only control. Keep multiple restore points, use a retention period that covers the time needed to discover a compromise, and preserve at least one copy outside the primary hosting account. A backup stored on the same server or under the same administrator account may be unavailable when the server or account is compromised.

3. Create a disaster-recovery workflow

  1. Declare the incident: record the time, symptoms, affected services, and person coordinating recovery. Avoid making untracked changes while investigating.
  2. Contain the cause: if compromise is suspected, restrict access, rotate exposed credentials when practical, preserve relevant logs, and separate affected systems from the clean recovery environment.
  3. Select the restore point: choose a copy that meets the RPO and predates the suspected corruption or unauthorized change. Document why it was selected.
  4. Prepare the destination: use a clean hosting account, isolated server, or temporary environment. Match supported runtime, database, storage, and permissions before importing data.
  5. Restore in dependency order: deploy files and configuration, create the database, import data, connect the application, restore media, and then re-enable scheduled tasks and integrations.
  6. Validate privately: test login, navigation, forms, search, uploads, checkout or membership functions, redirects, email, SSL, and administrator access before changing public DNS.
  7. Return to service: update DNS or routing only after the restored site is ready. Monitor errors, traffic, logs, and transactions during the recovery window.
  8. Review and improve: record actual RPO and RTO, failed steps, missing access, and any data that required manual reconciliation.

4. Reusable restore-test template

Copy this template into your operations documentation and complete it for each important site:

  • Site and owner: [domain, business owner, technical owner]
  • Systems in scope: [files, database, media, email, DNS, SSL, integrations]
  • Target RPO: [maximum acceptable data loss]
  • Target RTO: [maximum acceptable downtime]
  • Backup schedule: [frequency and backup type]
  • Retention: [daily, weekly, monthly or other restore points]
  • Storage locations: [primary, off-site, offline or otherwise isolated copy]
  • Access controls: [accounts, multifactor authentication, encryption, key ownership]
  • Test date and restore point: [date, timestamp, backup identifier]
  • Destination: [staging server, temporary host, or isolated environment]
  • Validation results: [application, database, media, email, DNS, SSL, integrations]
  • Rollback decision: [conditions for returning to the previous system]
  • Gaps and assigned actions: [owner and due date for each fix]

What to double-check

Before relying on a hosting backup feature, confirm exactly what it captures, how often it runs, how long copies are retained, and whether you can download or move them to independent storage. “Backup enabled” is not the same as “recovery tested.” Check whether the provider supports point-in-time database recovery, individual file restoration, full-account restoration, and recovery to a different server or hosting plan.

Protect the backup system as carefully as the website. Limit who can delete or alter copies, use separate administrator accounts, enable multifactor authentication where available, and avoid embedding credentials in scripts or shared documents. If backups are encrypted, document key ownership and the recovery process; encryption without recoverable keys can make a backup unusable.

For WordPress, test a restore with the same PHP and database compatibility requirements as production, then review plugin and theme behavior. Before a major version, PHP, plugin, or theme change, create a known restore point and confirm that rollback is possible. The PHP compatibility and rollback guide can be used alongside this process.

Also verify operational dependencies. A restored website may still appear broken if DNS points to the old server, an SSL certificate is missing, a payment key is invalid, cron jobs are disabled, or email records were not recreated. For broader hosting changes, see the website migration guide and the monthly uptime and reliability checklist.

Common mistakes

  • Keeping only one copy: one damaged, deleted, or encrypted copy is not a recovery strategy. Use more than one location and keep one copy independent of the production account.
  • Backing up files but not the database: this can restore the site’s appearance while losing content, users, orders, or settings.
  • Ignoring email and DNS: website files may be recoverable while customer communication and domain routing remain unavailable.
  • Testing only that a job completed: a successful backup task does not prove that the archive is complete or restorable.
  • Restoring over production first: an isolated test reduces the risk of overwriting evidence, current data, or a usable version.
  • Retaining copies for too little time: corruption or unauthorized access may not be noticed immediately. Align retention with the time needed to detect and investigate problems.
  • Leaving recovery knowledge with one person: document contacts, access methods, provider details, and the restore sequence so another authorized administrator can follow it.

When to revisit

Run a restore test at least quarterly for important websites, and sooner after changing hosting providers, backup tools, domains, CMS versions, runtime versions, or major integrations. Test before high-risk updates and before seasonal planning cycles when traffic, staffing, or business recovery targets may change.

Review the plan whenever the site’s data volume or change rate increases, when email moves providers, when a new administrator joins or leaves, or when an incident reveals a gap. Recheck retention, off-site storage, account permissions, encryption keys, and contact details at the same time.

For the next review, schedule a controlled restore, measure the actual recovery time, compare the recovered data with the stated RPO, and assign owners to every failed check. A short, documented test is more useful than an unverified promise that backups exist.

Related Topics

#website backups#disaster recovery#RPO#RTO#WordPress security#hosting security
S

Smart Hosting Editorial Team

Hosting and Security Editors

Senior editor and content strategist. Writing about technology, design, and the future of digital media. Follow along for deep dives into the industry's moving parts.