WordPress maintenance·8 min read·By the OmegaCloud team

WordPress maintenance on a VPS: 8 checks to run every month

Developer running WordPress maintenance commands over SSH on a VPS, with a monthly checklist open on a second screen

Every month, run through eight WordPress maintenance checks: restorable backups, updates, PHP and operating system, plugins and user accounts, file integrity, performance, Site Health and VPS resources. Book one slot per site, and take a snapshot before you change anything.

Most WordPress maintenance guides stop at the dashboard or point you to an agency retainer. Here, everything happens on the VPS side, on the command line, with native or open-source tools.

In brief

  • Security patches go on as soon as they are released, while major versions go through a staging site first.
  • A snapshot comes before every update, and an off-server backup completes it.
  • An 8-point checklist runs once a month, in a fixed slot.
  • After each intervention, a security and performance check confirms the site is holding up.

How often should you update WordPress?

The rule fits in three lines. Vulnerability published: update within 24 to 48 hours. Major version: wait a few days, then test on staging. Everything else: monthly cycle.

There are three cases to tell apart:

Plugins and themes expose a site far more than WordPress core does. Public vulnerability databases (NVD, CVE) show it with every release, which is why check 4 gives them a section of their own.

Snapshot before every update: yes, but not on its own

Take a snapshot before every update, without exception. But it does not replace a backup.

A snapshot is an image of the VPS disk. Rolling back takes a few minutes, which makes it the natural safety net for an update. The catch is that it stays with your host: if the account is compromised or the infrastructure has a problem, it disappears along with everything else.

A backup (files plus a database export) is copied off the server. Follow the 3-2-1 rule: three copies, on two different media, one of them off-site. Security guidance generally advises keeping at least one copy disconnected from the production system, which is precisely what protects you from ransomware. Just before you start work:

cd /var/www/mysite
wp db export /var/backups/mysite-$(date +%F).sql

With the OmegaCloud panel, snapshots and backups are managed from the same place. The database export is then run over SSH.

How to create a WordPress staging site on a VPS

The first setup takes time, then the procedure gets quicker with each repeat. Here is how to do it:

  1. Dedicated subdomain and vhost. Create staging.example.co.uk with its own Nginx or Apache block and its own directory.
  2. Copy the files and the database.
    rsync -a /var/www/mysite/ /var/www/staging/
    wp db export prod.sql --path=/var/www/mysite
    wp db import prod.sql --path=/var/www/staging
    
    Create a separate database first and enter its details in the staging wp-config.php.
  3. Replace the URLs.
    wp search-replace 'https://example.co.uk' 'https://staging.example.co.uk' --skip-columns=guid --dry-run --path=/var/www/staging
    
    Run it again without --dry-run once the output looks right.
  4. Locked access and no indexing. Add HTTP authentication and a blocking header in the Nginx vhost, then tick "Discourage search engines from indexing this site".
    auth_basic "Staging";
    auth_basic_user_file /etc/nginx/.htpasswd;
    add_header X-Robots-Tag "noindex, nofollow" always;
    
  5. No real emails or payments. Disable outgoing mail with a mu-plugin containing add_filter('pre_wp_mail', '__return_false');. On WooCommerce, switch the payment gateways to test mode.

If the site is heavy or business-critical, a staging copy on a second small VPS keeps resources separate: a load test no longer touches production.

Understand: testing on the live site means using your visitors as guinea pigs. A clash between two plugins shows up on staging as an error message. On production, it shows up as a white screen at peak traffic.

The monthly WordPress maintenance checklist: 8 checks to run

Block out a fixed slot, for example the first Tuesday of the month. You can copy this table straight into your tracking tool.

Check Command or tool Warning sign
1. Backups Date of the last off-server copy, restore to staging Copy too old or restore fails
2. Updates wp core check-update, wp plugin list --update=available Security fix pending
3. PHP and OS php -v, apt list --upgradable PHP out of support, kernel needs a reboot
4. Plugins and accounts wp user list --role=administrator Unknown account, abandoned plugin
5. Integrity and security wp core verify-checksums Modified file, certificate close to expiry
6. Performance TTFB, Core Web Vitals, wp db optimize Slower than the previous month
7. Site Health and cron Tools > Site Health, wp cron event list Overdue task, recurring PHP error
8. VPS resources df -h, free -h, panel monitoring Disk > 80%, swap in constant use

1. Backups: check they exist and can be restored

Check the date of the last off-server copy, then restore it to staging at least once a month. A backup that has never been restored is not a backup. The classic mistake is discovering a truncated export on the day of the incident.

2. Core, plugins and themes up to date

wp core check-update
wp plugin list --update=available
wp theme list

Apply updates on staging first (wp plugin update --all), then on production, after a snapshot.

3. PHP and operating system

WordPress.org recommends PHP 8.3 or higher. For end-of-support dates, the php.net calendar is the reference: PHP 8.1 has not been maintained since the end of 2025, and PHP 8.2 stops at the end of 2026. On the system side:

sudo apt update && sudo apt upgrade
sudo apt install unattended-upgrades
[ -f /var/run/reboot-required ] && echo "Reboot required"

unattended-upgrades applies security fixes on its own. Reboot if the kernel has changed.

4. Plugins and user accounts

Delete unused plugins and themes rather than just deactivating them. Spot any that have not been updated for more than a year: their author has probably moved on. Then review the accounts:

wp user list --role=administrator

Remove stale accounts and enforce strong passwords and 2FA. Our article on how to access the WordPress admin and secure it on a VPS covers this in detail.

5. Integrity and security

wp core verify-checksums
wp plugin verify-checksums --all

The second command only covers plugins published in the official directory. Then check file permissions (644 for files, 755 for directories, chmod 640 wp-config.php), SSH attempts and fail2ban, and the certificate expiry date:

sudo fail2ban-client status sshd
echo | openssl s_client -servername example.co.uk -connect example.co.uk:443 2>/dev/null | openssl x509 -noout -enddate

To harden the server itself, see how to secure an Ubuntu VPS for production.

6. Performance

Measure TTFB on three typical pages (home page, blog post, product page) and compare with the previous month:

curl -o /dev/null -s -w "TTFB: %{time_starttransfer}s\n" https://example.co.uk/ 

Add the Core Web Vitals, then clean up the database:

wp transient delete --expired
wp db optimize

Also check that OPcache and the object cache are responding. For tuning, see our tutorial on how to optimise a VPS for hosting WordPress.

7. Site Health, errors and cron

Open Tools > Site Health and deal with the critical items. Enable WP_DEBUG_LOG on staging only, never on production. Both constants are needed, since WP_DEBUG_LOG only takes effect when WP_DEBUG is on. Add this to the staging wp-config.php:

define('WP_DEBUG', true);
define('WP_DEBUG_LOG', true);
``` Then list the scheduled tasks:

```bash
wp cron event list

On a VPS, replace WP-Cron with a real system cron job. In wp-config.php: define('DISABLE_WP_CRON', true);, then in crontab -e:

*/5 * * * * cd /var/www/mysite && wp cron event run --due-now >/dev/null 2>&1

8. VPS resources

Look at disk space (df -h), RAM and swap (free -h), CPU load over 30 days in the panel monitoring, log size (journalctl --disk-usage, logrotate rules) and the space taken by snapshots. Rough thresholds: disk above 80%, swap in constant use. Our complete guide to optimising VPS performance takes this check further.

Checking security and performance after an update

Allow ten minutes. During an update, WordPress creates a .maintenance file in the site root and deletes it at the end: the built-in maintenance mode only lasts for the duration of the update. If the maintenance page is still showing after an interrupted update, delete that file.

  1. Click through the critical paths: home page, contact form, checkout, admin login.
  2. Check the HTTP status codes: curl -I https://example.co.uk/.
  3. Read the PHP and Nginx error logs for the last hour.
  4. Compare TTFB with the measurement from before.
  5. Run wp core verify-checksums again.

If something has regressed, restore the snapshot, then analyse the problem on staging, not on the live site.

A VPS built to make this routine easier

This routine mostly depends on having the right tools to hand. On OmegaCloud's WordPress VPS plan, snapshots and backups are managed from the panel. The built-in monitoring serves check 8, and the emergency console rescues you when an SSH change goes wrong. KVM virtualisation without resource over-allocation keeps your measurements comparable from one month to the next. To see why, read how a VPS shares a physical server through its hypervisor. Hosting is in Paris, GDPR-compliant, with Anti-DDoS included.

Plans are currently available for pre-order, with availability announced for 1 November 2026.

Frequently asked questions

How often should you update WordPress and its plugins?

Apply security fixes within 24 to 48 hours of release. Major versions wait a few days and go through staging. The rest (plugins, themes) is handled in the monthly cycle.

Should you take a snapshot before every WordPress update?

Yes, it is the fastest way back. Pair it with a wp db export and a backup copied off the server.

What is the difference between a snapshot and a WordPress backup?

A snapshot is an image of the VPS disk, restorable in a few minutes but stored with your host. A backup (files plus a database export) is copied off the server. The first undoes a failed update, the second lets you survive a more serious incident.

How do you create a WordPress staging environment on a VPS?

Create a subdomain with its own vhost, copy the files with rsync and the database with wp db export then wp db import. Fix the URLs with wp search-replace, password-protect access, block indexing and switch off emails and payments.

Should you enable automatic plugin updates in WordPress?

Only for a few critical, well-maintained plugins. The others stay in the monthly cycle, with a test on staging. Automatic security updates for core, on the other hand, stay on by default.

What should you do if a WordPress update breaks the site?

Restore the snapshot taken before the intervention, then delete .maintenance if the maintenance page persists. Next, repeat the update on staging and read the PHP error log to isolate the plugin responsible.

To put this routine in place, start with a single site: schedule the monthly slot, build the staging copy and test a full restore. Once the 8 checks fit into under an hour, you can repeat them across your other sites, on a base such as OmegaCloud's WordPress VPS plan, with no minimum commitment period.

Related articles

Hosting

Navigation