Complete restoration requires database and files from the same point in time
WordPress documentation says both database and files are needed to completely restore a typical site. Copying the WordPress directory on the server does not normally include the separate MySQL or MariaDB database. SQL alone cannot restore themes, plugins, uploads, wp-config.php, or .htaccess.
Check consistency before restoring only part of a site
Orders, members, reservations, inventory, and forms can change both database and uploads. Mixing an old database with newer files can break references and state. Record capture time, site version, WordPress, PHP and database versions, plugins, hashes, and exclusions in the backup set.
Scroll horizontally to compare backup scope, commonly missed items, and restore checks.
| Category | Main scope | Often missed | Restore checks |
|---|---|---|---|
| Database | Posts, pages, comments, users, settings, and tables stored by plugins and themes | Other prefixes, custom tables, external databases, tables without primary keys, and orders or jobs in progress | Check all tables, encoding, row counts, URLs, login, and business records |
| WordPress files | wp-content, uploads, plugins, themes, mu-plugins, wp-config.php, .htaccess, and additional code | Uploads outside WordPress, symbolic links, exclusions, and large files | Check ownership and permissions, missing images, themes, plugins, secrets, and generated files |
| Outside WordPress | DNS, domains, TLS, email, CDN, WAF, object storage, cron, environment variables, and external SaaS | Configuration, API keys, webhooks, billing, and contracts not included in plugin backups | Keep separate inventories and exports; check recovery order, contacts, and reissue procedures |
Set frequency and recovery methods from RPO, RTO, and change volume
Do not choose weekly or daily uniformly; work backward from acceptable data loss and business downtime. WordPress also advises more frequent backups for active sites and database backups before updates.
RPO: how far back can you restore?
Daily backups may lose up to about 24 hours of data. For frequent orders, reservations, or membership changes, consider shorter intervals, real-time methods, and preserving subsequent changes.
RTO: how long to restore operations?
Measure time for cloud download, extraction, database import, DNS, TLS, cache, testing, and approval. Do not estimate from a one-click label alone.
Backup triggers
In addition to scheduled backups, create restore points before WordPress, PHP, theme, or plugin updates; migration; URL changes; database work; and bulk editing.
Retention and generations
Keep generations extending beyond the time it may take to notice an incident or compromise. Set retention and disposal according to capacity, law, personal data, and deletion requests.
Prepare storage and permissions that cannot be lost with production
WordPress recommends keeping several recent generations in different locations and manually checking as well as automating backups. A same-server copy is fast but can be lost with production through server failure, compromise, deletion, or contract termination.
Scroll horizontally to compare roles, shared-loss examples, and required checks.
| Destination | Role | Examples of shared loss | Checks |
|---|---|---|---|
| Same server | Quick rollback of recent changes or temporary staging during creation | Disk failure, compromise, encryption, exhausted capacity, contract termination, or the same delete operation | Do not use as the only final destination; check that it is outside the public directory, capacity monitoring, and automatic deletion |
| Host snapshots | Fast recovery of a server or volume | Shared contract or admin console, limited retention, capture timing, or waiting for a restore request | Check scope, frequency, retention, separate regions, retrieval, and recovery time in the contract |
| Cloud or object storage in a separate account | Automated off-production storage, generations, and remote copies | Shared admin credentials, synchronized deletion, public settings, lost keys, or account suspension | Test least privilege, MFA, encryption, retention locks, audit logs, decryption, and retrieval |
| Local or offline | Storage independent of cloud outages or account compromise | Lost devices, failed media, stale copies, or lost encryption keys | Check encryption, separate location, regular connection, media replacement, separately stored keys, and restore readability |
Compare five options by recovery method and storage, not rank
This is a snapshot of WordPress.org descriptions checked on July 25, 2026. Do not treat vendor ratings or counts as proof of reliability. Compare free and paid boundaries, storage, recovery when WordPress is down, and exclusions with the same tests.
Scroll horizontally to compare use, free and contract terms, storage, and recovery considerations.
| Option | Suitable workflow | Free and contract terms | Storage | Recovery considerations |
|---|---|---|---|---|
| UpdraftPlus | Separate database, plugins, themes, and uploads; run manual or scheduled backups and restore through the admin interface. | Test manual and scheduled backups, several major remote destinations, component-based restore, and migration in the free edition. Incremental backups, automatic pre-update backups, multiple destinations, and database encryption are Premium features. | Free and Premium destinations differ; check required destinations and the contract scope for sending to multiple locations. | Test whether components represent the same point in time, all archives are present, and recovery works when the admin interface is unavailable. |
| BackWPup | Configure WHAT, WHEN, and WHERE for scope, frequency, and multiple destinations, then manage history and logs. | Test database and file backups, scheduling, multiple cloud destinations, and restoration of manually uploaded backups in Free. Direct cloud restore, encryption, migration, and the standalone restore app are Pro features. | Check multiple destinations for one job and retention by destination. Some combinations, including Google Drive, require Pro. | Restore-source operations differ between Free and Pro. Review restoration and security fixes in release notes before installation. |
| Jetpack VaultPress Backup | External cloud storage, activity logs, daily or real-time backups, and restoration while the site is down are required. | A paid Jetpack plan including Backup is required. Check frequency, retention, storage, and WooCommerce support for each plan before subscribing. | Stored in Jetpack's cloud rather than directly in Google Drive or Dropbox; check download availability. | The official FAQ says Multisite and split site/home URLs are unsupported. Test credentials, offline restoration, and order-retention conditions. |
| Duplicator Lite | Package database and files together for cloning, staging, and host or domain migration. | Test manual packages and migration in Lite. Scheduled backups, cloud integrations, Recovery Points, and Multisite are among Pro features. | Local downloads are the baseline; check the Pro contract scope for cloud, FTP/SFTP, and other destinations. | Test execution time and server resources for a large site, installer protection, and restoration into an empty environment. |
| All-in-One WP Migration | Export the database, media, themes, and plugins to a .wpress file and import into another WordPress site for migration. | Test basic export and import. Unlimited size, cloud, Multisite, scheduling, and database filtering require Premium Extensions. | Check local export separately from the cost, authentication, and retention of required cloud extensions. | Test host upload limits, PHP and disk capacity, retrieval of the .wpress file, and cache and permalinks after import. |
Configure backup, verification, and notification as one job
Do not stop after the first full backup. Include scheduling, retention, monitoring, and restoration ownership in operations. For live sales or bookings, verify the consistency method knowing state can change during capture.
- 1
Record scope and exclusions
List all database tables, WordPress files, additional directories, and external data. Exclude only cache and temporary files, and review to avoid accidentally excluding custom tables or uploads.
- 2
Set frequency, retention, and destination from RPO
Define the roles of full, incremental, and real-time backups, run times, generations, capacity limits, failure cleanup, and at least one off-production copy.
- 3
Separate permissions and secrets
Use dedicated credentials, least privilege, MFA, encryption, and audit logs for storage. Keep decryption keys and recovery procedures on a separate path that will survive the same incident.
- 4
Monitor success, failure, and missed runs
Notify the last successful time, scope, size, duration, destination, hash, and failure reason. Treat missing success notifications as failures too.
- 5
Create a pre-update restore point
Take a manual backup before changes to WordPress, PHP, themes, plugins, databases, or bulk edits; verify completion and off-site storage before proceeding.
Connect backup scope and settings to recovery requirements
Official WordPress documentation explains that database and files form one backup set. The BackWPup screenshots shown at the verification date illustrate configuring scope, frequency, and storage separately. Interfaces change, so reread scope and destinations in the version you install.


A backup passes only after restoration in another environment
Opening an archive, matching file counts, or reading SQL are intermediate checks. Restore into an empty environment or staging isolated from production and exercise user and business operations. Manage personal and secret information; do not hand production backups directly to external testers.
Preparation and evidence
- Record restore date, RPO, WordPress, PHP and database versions, URL, complete backup set, and hashes.
- Prepare an empty environment, database, capacity, permissions, TLS, decryption keys, and access to retrieve the backup.
- Save start and end times, owner, operation logs, errors, and manual adjustments.
Restoration order and consistency
- Follow WordPress's basic order: restore files, then import the database.
- Check wp-config.php database connection, URL, encoding, ownership and permissions, cron, and cache.
- Reconcile images and attachments, custom tables, orders, reservations, members, and external object references.
User and business operations
- Check home, articles, search, 404, images, login, administration, and actions for each role.
- Safely test relevant business operations such as forms, email, webhooks, purchases, refunds, bookings, and member updates.
- Check DNS, TLS, CDN, WAF, and external APIs without accidentally sending to production.
Acceptance and improvement
- Measured RPO and RTO meet targets and the owner can restore using the procedure alone.
- Fix missing data, manual adjustments, insufficient permissions or capacity, and old keys as findings.
- Back up and restore again, repeating until no findings remain.
During an incident, protect evidence and new data before restoring
During compromise, mistakes, failed updates, or server outages, hastily overwriting production can lose evidence or orders created after the backup. Decide impact, restore point, subsequent data, and communication order before recovery.
- 1
Stop changes and further harm
Stop production changes, scheduled jobs, synchronized deletion, and compromised credentials. Preserve times, symptoms, URLs, logs, screens, and configuration, using a clean device and communication channel.
- 2
Agree on a restore point and data at risk
Check the last known good point and backup success, location, and integrity. Determine whether later orders, reservations, memberships, posts, and form data can be extracted or re-entered.
- 3
Restore in isolation and remove the cause
Restore database and files from the same set. For compromise, remove vulnerabilities, accounts, secret keys, and backdoors. Do not return an unresolved cause to production.
- 4
Switch after acceptance
Check key URLs and business operations, monitoring, TLS, email, and webhooks; switch DNS or traffic after approval. Isolate the old environment according to evidence-retention policy.
- 5
Apply later data and monitor
Apply legitimate post-restore-point data without duplication or loss and monitor 404s, 5xx, authentication, submissions, orders, cron, and backup resumption.
Before and after launch checklist
Record scope, time, destination, hash, notification, restore result, RPO, RTO, and owner together, and recheck after updates or contract and permission changes.
Before design and purchase
- Explain included and excluded database, file, and external WordPress components.
- Define RPO, RTO, frequency, retention, capacity, storage, and disposal.
- Check whether required features are free, paid, or extensions.
Routine operations
- Monitor the last success, missed runs, sudden size changes, and destination failures.
- Maintain an off-production copy, MFA, least privilege, encryption, and separately stored keys.
- Create a manual restore point before updates, migration, or database work.
Periodic tests
- Retrieve, decrypt, and restore database and files in a separate environment.
- Test user, administrator, and business operations and measured RPO and RTO.
- Fix findings, back up and restore again, and update procedures and owners.
Primary sources consulted
Scope and restore order were checked against WordPress documentation and candidate features and contract boundaries against WordPress.org plugin pages as of July 25, 2026. Versions, prices, and destinations are not fixed here; recheck them when installing.
- WordPress Developer Resources: Backups
- WordPress Developer Resources: Backing Up Your WordPress Files
- WordPress Developer Resources: Backing Up Your Database
- WordPress.org plugin directory: UpdraftPlus
- WordPress.org plugin directory: BackWPup
- WordPress.org plugin directory: Jetpack VaultPress Backup
- WordPress.org plugin directory: Duplicator
- WordPress.org plugin directory: All-in-One WP Migration
Frequently asked questions
Are a host's automatic backups enough?
Judge whether scope, frequency, retention, shared failure risks, retrieval, and recovery time meet your RPO and RTO. Do not rely solely on one contract or admin interface; also maintain an off-production copy you can retrieve and restore yourself.
Does a backup success email prove I can restore?
No. Check the complete database and file set, scope, size, hash, and destination, then restore in another environment and verify key operations and RPO and RTO.
Is a daily database-only backup enough?
It can supplement backups when posts change but files rarely do; complete restoration normally requires files too. Capture uploads, themes, plugins, wp-config.php, .htaccess, and additional code as a same-point-in-time set.
Can I keep a backup on the same server?
It can provide a quick rollback but must not be the only copy. Disk failure, compromise, deletion, exhausted capacity, or account termination can remove it with production, so add independently managed storage such as another account or an offline copy.
Is restoring the latest backup safe after a compromise?
The latest backup may also contain the compromise. Investigate when and how it occurred, restore a known-clean point in isolation, and remove vulnerabilities, accounts, secret keys, and backdoors before switching production.
Related WordPress operations and recovery guides
Completion means safe recovery within the target time, not just a successful backup
Inventory the database, files, and external configuration, then set frequency, retention, and destinations from RPO and RTO. Provide a copy that cannot be lost with production, dedicated permissions, encryption, separately stored keys, and failure monitoring. Compare free and paid boundaries and recovery while the site is down, then restore a same-point-in-time database and files to another environment. Make passing critical operations and measured RPO and RTO, fixing findings, and retesting a release condition.
Editorial and verification policy
Finite Field Editorial Team
We compare WordPress.org primary sources without converting vendor claims into recovery guarantees, and organize a process for assessing scope, storage independence, free and paid boundaries, measured RPO and RTO, and restoration tests.
For restore failures, compromise, or migration decisions
Discuss WordPress backup design and recovery
Prepare the affected URL, incident time, last known good point, backup inventory, database, files and external configuration, destinations, logs, RPO and RTO, recent changes, and subsequent orders, reservations, and form submissions.
Discuss backup design and recoveryWe will review the situation and available data before advising whether we can help. We do not guarantee zero downtime, complete recovery, zero data loss, recovery within a specified time, or removal of compromise in advance.
