If a WordPress change saves but the live site still shows the old version, first determine whether the problem is visibility or installation. A browser, WordPress cache plugin, hosting cache, reverse proxy, CDN, incorrect template, or wrong URL can serve an older copy. If the update notice itself is missing, WordPress may simply not have reached its next randomized update check. If an update failed, permissions, connectivity, a stale .maintenance file, or a fatal plugin/theme error are more likely. Work through the checks below in order, beginning with a private-window test and ending with Site Health, filesystem checks, or Recovery Mode.
Antihost offers managed WordPress hosting with free migration, one-click staging and a 14-day free trial.
Identify what “not showing” means
There are three different failures that are often described with the same sentence. Separating them prevents you from deleting caches when the update was never installed, or editing files when visitors are only seeing an old response.
As an Amazon Associate I earn from qualifying purchases.
Free tools Windows power users keep installed
One-click scans. No signup required.
| What you observe | Most likely area | First check |
|---|---|---|
| Your editor confirms a save, but logged-out visitors see old text | Browser, page cache, host cache, reverse proxy, or CDN | Open the exact public URL in a private window, then purge caches that actually exist |
| The page is unchanged even in a private window and while logged in | Wrong page, homepage assignment, template, theme, upload, or source | Verify the public address and the template or source that renders it |
| A newly released core, plugin, or theme update is absent from Dashboard > Updates | Randomized update-check timing, blocked WordPress.org requests, or a site-wide connectivity problem | Refresh later and inspect Tools > Site Health |
| An update starts and reports a download or installation failure | HTTP connectivity, filesystem ownership, permissions, disk or directory write access | Back up files and database, read the exact error, and check Site Health and permissions |
| The site shows a fatal PHP error after updating | Incompatible or damaged plugin or theme | Use WordPress Recovery Mode from the administrator email |
| Automatic updates never run while manual updates work | WordPress Cron, host policy, disabled auto-updates, or blocked background requests | Review Site Health and the automatic-update settings |
WordPress does not include a page-cache system by default. Caching exists only if a plugin, hosting platform, web server, reverse proxy, CDN, or browser has introduced it. That distinction matters: installing a cache plugin just to clear a cache that is not present can add another variable without fixing the original problem.
Follow this safe troubleshooting sequence
1. Confirm that the change was actually saved
Look for the editor’s success notice, not merely a button press or a disappearing spinner. Reopen the item in the editor and confirm that the new text, image, setting, or code is still present. Then visit the exact public URL, including the correct path, language version, or preview domain.
#1 Best Overall
Check the result while logged out or in a private browser window. Administrators can receive a fresh or privileged response while ordinary visitors receive a cached page. If the private view contains the new content, the save succeeded and the next task is cache invalidation. If both views are old, continue to the page, template, and source checks rather than repeatedly saving.
2. Purge every cache layer that is really enabled
Clear the cache in any WordPress performance or caching plugin that your site uses. Next, purge a hosting or server cache if your host provides one, and purge a CDN or reverse-proxy cache if traffic passes through one. These layers are independent: clearing the browser cannot remove an object stored at the host, and clearing a plugin cannot remove an edge copy held by a CDN.
Managed hosts may cache complete pages at the server level. A reverse proxy can continue serving the previous response until its expiration time or an explicit purge. After each purge, reload the public URL in a private window and, if possible, test from a second network. Do not assume that a hard reload proves every layer is clear; it only controls the browser making that request.
3. Verify the page, template, and source
WordPress can save a perfectly valid change in a location that is not used to render the page you are viewing. Compare the editor’s URL with the public address and confirm that you are editing the assigned homepage or the intended page in a multisite installation. Check which theme template produces the visible section, especially when a block theme, child theme, or custom template is active.
For images and other uploaded files, verify the upload location and the URL embedded in the page. For code or configuration, verify the source file or setting that the active theme and environment actually load. The WordPress troubleshooting FAQ recommends checking the address, template, upload, and page source when a change appears to have no effect. View the page source or browser inspector only to establish which resource is being delivered; do not edit generated HTML as a substitute for changing the source.
4. Allow for WordPress’s update-check schedule
A release can be available on WordPress.org while your Dashboard still shows no notice. WordPress’s troubleshooting FAQ states: “Your blog is programmed to check for updates every 12 hours, but the timing of that check is purely random.” The approximately 12-hour interval is a scheduled check, not a promise that every site will display a notice at the same clock time.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsOpen Dashboard > Updates and check again later. If the notice remains absent after a reasonable interval, treat it as a connectivity or site-health problem rather than an ordinary delay. Tools > Site Health can show whether the site can reach WordPress.org and whether HTTP requests or background tasks are being blocked.
5. Back up before retrying a failed update
Before you retry an update that reported an error, make a current backup of the database and WordPress files. A failed package download is usually recoverable, but an interrupted file replacement or incompatible extension can leave a site partially updated. Keep the backup outside the same hosting account when possible so that a hosting failure does not remove both copies.
Read the dashboard message exactly. WordPress documents separate download failures, installation failures, and connection errors; each points to a different layer. A download error suggests WordPress.org access, DNS, TLS, or an HTTP request problem. An installation error commonly points to write access, ownership, a locked file, or an incomplete package. Do not infer a permissions fix from a generic “update failed” notice until you have checked the detailed error and Site Health.
6. Remove a stale .maintenance file only when WordPress left one behind
During an update, WordPress places a .maintenance file in the site’s root directory and removes it when the process finishes. If the process stops unexpectedly, the file can remain and make the site report that it is briefly unavailable even after the update attempt has ended.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Use the documented FTP or hosting file-manager procedure to inspect the root directory. If a failed update left the root .maintenance file in place, remove that file, reload the site, and then follow the manual-update or retry instructions appropriate to the reported error. Do not delete unrelated files, and do not treat removal as a cure for a package that never downloaded or a directory that is not writable.
7. Inspect Site Health and automatic-update prerequisites
Go to Tools > Site Health and read both the Status and Info areas. Pay particular attention to warnings such as “Could not reach WordPress.org,” blocked HTTP requests, background-update failures, and filesystem permissions indicating that WordPress cannot write to required directories. These messages provide evidence that is more useful than repeatedly pressing Update.
Plugin and theme automatic updates depend on WordPress Cron. When scheduled updates do not run, Site Health can reveal Cron symptoms or other background-task failures. Automatic-update controls were introduced in WordPress 5.5, and a hosting provider or another plugin can disable them. A site can therefore have a current WordPress version while its plugins and themes remain unchanged because the scheduled mechanism is disabled or unable to execute.
Use visibility to choose the next diagnostic
Only logged-out visitors see the old page
This pattern strongly favors a cache outside the editor session. Test a private window, then clear the WordPress cache plugin, hosting cache, and CDN or reverse-proxy cache in that order if those layers are enabled. Check again after each purge so you know which layer was responsible. If the old response persists, ask the host whether another page-cache tier or edge rule is active.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteEvery browser, including a private window, shows the old page
When no browser sees the change, confirm that the editor saved the expected object and that the public URL maps to it. Check the assigned homepage, the active theme and template, and the upload or source location. A page builder, block template, or child theme can render content from a different location than the screen you edited. Page source can help you identify the template output or asset URL being delivered.
Rank #4
The editor itself reverts to the old value
This is not a display-cache symptom. The save may have failed, a permission rule may have rejected the write, or another process may have overwritten the value. Reopen the editor to see whether the new value persisted, note any error notice, and use Site Health’s filesystem and HTTP findings before attempting another update. Preserve a backup if the change affects code, a theme, or a plugin.
The whole Updates screen is missing notices
First allow for the randomized check interval. Then inspect Tools > Site Health for WordPress.org reachability, blocked HTTP requests, and background-update failures. If core, plugin, theme, and translation notices all disappear together, a site-wide request or Cron problem is more plausible than four independent release delays.
Diagnose update failures by the evidence shown
| Evidence | What it indicates | Action |
|---|---|---|
| “Could not reach WordPress.org” or a blocked HTTP request | The site cannot retrieve update metadata or packages | Resolve outbound HTTP, DNS, TLS, firewall, or host restrictions, then check Dashboard > Updates again |
| Download failure | The package was not retrieved successfully | Check connectivity and available storage; retry only after the underlying error is addressed |
| Installation failure with write or permission wording | WordPress cannot replace or create required files | Have the host or administrator correct ownership and directory permissions, then retry from a backup |
| Site remains in maintenance mode after an interrupted update | The root .maintenance marker was not removed |
Remove only that documented root file through FTP or the host file manager |
| Fatal PHP error immediately after activation | A plugin or theme is incompatible, damaged, or calling unavailable code | Use Recovery Mode to pause the identified component and repair or replace it |
| No manual error, but scheduled updates never run | Cron or automatic-update controls are not functioning | Review Site Health, host policy, and auto-update settings |
Fix filesystem and permission problems safely
WordPress must be able to write to the directories and files involved in an update. Ownership can be wrong even when a directory appears writable in a file manager, particularly after a migration, a server-user change, or a manual upload. The Updating WordPress documentation identifies filesystem ownership and permissions as common causes of update failures.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →- Record the exact file or directory named in the error.
- Ask the host which system user runs PHP and whether that user owns or can write to the WordPress files.
- Correct ownership and permissions using the host’s supported method rather than applying broad write access to the entire account.
- Retry only after confirming that the database and files are backed up.
- Recheck Tools > Site Health to verify that the warning has cleared.
A permission repair should not be used to conceal an HTTP download failure, and a successful download does not prove that installation can write files. Treat retrieval and installation as separate tests.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Use Recovery Mode after a fatal plugin or theme error
WordPress Recovery Mode was introduced in WordPress 5.2. When a fatal PHP error is detected during a normal page load, WordPress can send the administrator an email containing a Recovery Mode link. WordPress describes it as a built-in feature that can help restore access without FTP or direct code editing.
- Open the Recovery Mode link in the administrator email while it is valid.
- Read the error details and identify the plugin or theme that WordPress has paused for your administrator session.
- Deactivate, repair, update, or replace that component from the dashboard, depending on the error and the component’s role.
- Exit Recovery Mode and load the site in a normal browser session.
- Test the affected pages, the dashboard, and any dependent functionality before reactivating a repaired component.
Recovery Mode is session-specific: it pauses the faulty component for the administrator so you can work, rather than silently fixing the underlying code. It also does not diagnose failures that occur only in Cron or other background tasks. For those, use Site Health and the server’s error logs, because a normal page load may never trigger the failing job.
Understand automatic updates and Cron
Manual and automatic updates use different paths. A manual update initiated from Dashboard > Updates can work while a scheduled update fails because the latter relies on WordPress Cron and background HTTP requests. Conversely, a site can show a scheduled task warning even when a manually started update succeeds.
Recommended Free Tools
Check whether plugin and theme auto-update controls are available and enabled. Those controls arrived in WordPress 5.5, but a host or another plugin can disable them. In Site Health, look for Cron failures, loopback or HTTP-request warnings, and filesystem messages. If the host blocks scheduled requests, ask for its supported Cron configuration rather than installing overlapping schedulers without understanding which one is active.
What not to do
- Do not keep clicking Update while an earlier process is still running; first read the status and check for a maintenance marker.
- Do not purge a cache that does not exist or install a new cache plugin solely to clear an unknown layer.
- Do not assume a missing notice means that a release is unavailable; the update check is randomized.
- Do not delete the entire WordPress directory when a single stale
.maintenancefile is documented as the issue. - Do not change broad filesystem permissions without understanding ownership and the host’s security model.
- Do not reactivate a plugin or theme that Recovery Mode paused until the fatal error has been addressed.
- Do not treat a normal front-end page load as proof that Cron and background updates work.
Prevent the next “not showing” incident
- Keep a recent database and file backup before core, plugin, or theme updates.
- Document which cache layers your site actually uses and where each purge control is located.
- After publishing an important change, verify the public URL in a private window rather than relying only on the editor preview.
- Review Tools > Site Health periodically so WordPress.org connectivity, Cron, loopback requests, and filesystem warnings are visible before an update is urgent.
- Keep administrator contact details current so Recovery Mode emails arrive when a fatal error pauses a component.
- Record the active theme, template, homepage assignment, and upload locations for anyone else who maintains the site.
When to involve your host or developer
Escalate when Site Health reports blocked outbound HTTP, the host controls a page cache you cannot purge, filesystem ownership cannot be changed from your account, or a fatal error remains after the component is paused. Include the exact dashboard message, the affected URL, the time of the attempt, whether logged-in and private-window views differ, and the Site Health warning text. That evidence lets support distinguish a cache purge from a server-permission or connectivity change.
For a site that is still serving visitors, preserve the backup and avoid speculative file edits while support investigates. If the update is security-critical, ask whether the host can apply the supported manual or managed update procedure after confirming a rollback path.
The Bottom Line
Most WordPress changes that save but do not appear are being hidden by a cache or rendered from a different page or template. Missing update notices can be normal for roughly 12 hours because checks are randomized. Failed updates require evidence-led checks of WordPress.org connectivity, Cron, filesystem ownership, permissions, and any leftover .maintenance file; a fatal error after activation calls for Recovery Mode.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
