Use the server-resource group to spot capacity symptoms that may affect the billing application. First establish which system the measurements describe.

Review the measurements

1. Open PD Modules → Guard → Report → Server.
2. Review CPU load, Memory (RAM) and Free disk space.
3. Compare the report time with the period when the problem occurred.
4. Refresh if you need a new point-in-time check.

CPU load is a load-average reading, not a measured CPU-usage percentage. RAM information may describe the whole node; it is not necessarily your hosting account’s allowance or PHP memory limit. Disk information is for the filesystem containing the installation, not a complete backup-storage inventory or customer quota report.

Separate unavailable from healthy

Restricted hosts may not expose the operating-system measurements. Unavailable or unknown does not mean zero usage. Check your hosting control panel’s account limits and resource history alongside the report.

The module uses fixed thresholds to grade the readings. A green result is not a forecast of capacity for the next busy period. A later quiet reading does not explain an earlier intermittent overload.

Investigate the capacity symptom

Your hosting team should compare the symptom and time window with the relevant system’s historical usage and your hosting account’s own limits. Use your normal administration tools or the infrastructure provider responsible for that server. Do not delete backups or unrelated files merely to clear a warning. A quiet reading now does not explain an earlier overload, and a node-wide measurement does not establish a particular customer’s usage.

See the PD Guard Blesta guide for related tasks, or explore PD Guard.