Open sourceIn the free reqad package (GPL-3.0).
Needs attention
The Needs attention tile on the dashboard counts what is wrong on the server or in a hosting account and shows the most serious problems first; Show all opens the full list. Problems are things like a certificate about to expire, mail DNS records that are missing, a site on a PHP version that gets no more security fixes, or a service that stopped. Each problem says which account it belongs to and links to the page where you fix it.
What is checked
In every hosting account:
| Check | Reported when |
|---|---|
| SSL certificates | A main or addon domain has no certificate, a self-signed one, one that expired, or one that expires within 14 days. Let's Encrypt certificates renew 30 days ahead, so 14 days left means the renewal is failing. The server's own hostname is checked too. |
| Mail DNS | For accounts with email: no MX record (critical - no mail arrives), or SPF, DKIM or DMARC missing, or a DKIM record that does not match the key the server signs with. The records are asked at the domain's own name server. Check & fix opens the details and can add the missing records at your DNS provider. |
| PHP versions | A site (main or addon domain) runs a PHP version whose security support has ended, or ends within 90 days. The default PHP version for new accounts is checked as well. The message lists the supported versions installed on the server. |
| Disk quota | An account with a quota uses 90% of it (warning) or all of it (critical). |
| WordPress | A site runs an older WordPress than the latest release (checked against wordpress.org once a day), or has plugin updates waiting as of the last time the site was opened in the WordPress Toolkit. |
| Infected filesPremium | The malware scanner found infected files in the account (critical). The item opens the list of that account's files. |
| Resource limits | A site ran out of PHP workers 3 or more separate times in the last 24 hours, or the session memory limit stopped a process (see Resource usage). |
On the server:
| Check | Reported when |
|---|---|
| Disk space | /, /home or /var is 90% full (critical at 95%). |
| Resource usage checks | The check of the accounts' CPU, memory and processes has not run for 15 minutes, or never ran because /etc/cron.d/reqad has no line for it. |
| Malware scannerPremium | ImunifyAV is installed but its service is not running, so no scans run. |
| Services | A service listed on the Services page is installed but not running, or has failed. |
| Nightly backup | No nightly remote backup is set up (info), the last run had failures, or no backup has finished for two days (critical). |
| Mail blocklists | The server's IP address is listed on Spamhaus ZEN, SpamCop or Barracuda, so other servers may reject its mail. Lists that refuse to answer the server's DNS resolver are skipped rather than reported. |
| Mail queue | 200 or more messages are waiting (critical at 1,000) - usually a remote server refusing mail, or an account sending spam. |
| Webmail and phpMyAdmin | Roundcube or phpMyAdmin is older than its latest release. The message shows the command that updates it. |
| Symbolic link protection | Warns when nginx follows links to other accounts' files (the protection is off), and when Apache refuses a .htaccess that asks for FollowSymLinks: that folder answers 500 until setup_symlink_protection.sh --htaccess changes it to SymLinksIfOwnerMatch. |
| Premium updatesPremium | The premium repository is switched on, but refuses this server: its Premium subscription or trial has ended, or its IP address is not marked Premium on your account page. The add-ons keep running without updates. Critical while that also stops every other update (servers with reqad-repo older than 1.0.4); the message shows the command that switches the premium repository off. |
| Reboot and Reqad updates | Listed on the command line; on the dashboard these keep their own banners with the Reboot and Update buttons. |
Severity
| Level | Meaning |
|---|---|
| Critical | Something is broken now or about to break: an expired certificate, no MX record, a stopped service, a full disk, no recent backup. |
| Warning | It works, but it is a risk: a self-signed certificate, missing SPF/DKIM/DMARC, an end-of-life PHP version, outdated software. |
| Info | Worth knowing: support for a PHP version ends soon, plugin updates are waiting, no nightly backup is set up. |
When one check finds three or more problems of the same level - several sites on an old PHP version, say - the list shows them as one row with the accounts involved. Click it to see each one.
Hiding a problem
Some problems are expected: a test site with a self-signed certificate, or a domain whose mail is handled elsewhere. Hide on a row in the full list takes it off the dashboard for 7 days, for 30 days, or until you show it again. If it is still a problem when the time is up, it comes back. Hidden problems are counted at the bottom of the list, where you can show them and bring one back with Show again.
A problem that is fixed disappears by itself the next time the checks run.
On the command line
reqad health list # what the last run found
reqad health list --account=bob # one account ('-' = the server itself)
reqad health run # check now
reqad health dismiss ssl:bob:bob.example --days=30
reqad health restore ssl:bob:bob.examplereqad health list --json gives the same list for scripts and monitoring. See Command line (CLI) for every option.
Need a hand? Create a free account or contact us.