WPHammer
Log in
  • How collection works
  • Viewing health data
  • Alerts and thresholds
  • Related

    • Server Health

    WPHammer monitors server health by collecting system metrics over SSH on a schedule. These metrics help you spot resource pressure, failing services, and degraded performance before they affect your WordPress sites.

    What gets collected

    Health checks collect six categories of metrics, each with its own threshold evaluation:

    CPU load

    The system load average, indicating how many processes are waiting for CPU time. High sustained load suggests the server is overworked — often caused by traffic spikes, runaway cron jobs, or under-provisioned resources.

    Memory usage

    The percentage of RAM in use. WPHammer distinguishes between active memory usage and cache/buffers, so the reading reflects actual memory pressure rather than the kernel's opportunistic caching.

    Swap usage

    How much swap space is being consumed. Any significant swap usage typically indicates the server has exhausted physical memory and is paging to disk — a strong performance warning.

    Disk usage

    The percentage of the root filesystem in use. Disk pressure can cause database writes to fail, log rotation to stall, and backups to break. See Disk Usage for the detailed monitoring page.

    PHP-FPM status

    Whether the PHP-FPM process manager is running and responsive. If PHP-FPM is down, WordPress sites on the server will return 502 errors.

    Zombie processes

    The count of zombie (defunct) processes on the server. A small number is normal, but a growing count suggests a process management issue that should be investigated.

    How collection works

    Health metrics are collected via SSH commands executed against the server. The collection runs on a schedule as a background job and stores each metric reading with a timestamp. This builds a historical record so you can review trends over time, not just the latest snapshot.

    Each metric reading is evaluated against configurable thresholds and assigned a status:

    • Ok — the value is within normal operating range
    • Warning — the value has crossed a cautionary threshold and should be monitored
    • Critical — the value indicates an immediate problem that needs attention

    Viewing health data

    Server health metrics are available on the Health tab of the server detail page. The view shows:

    • Current status for each metric type with color-coded indicators
    • The most recent value and its unit (percentage, count, etc.)
    • Historical readings to spot trends and recurring patterns

    The dashboard Health tab aggregates health data across all servers, so you can quickly identify which servers in your fleet need attention.

    Alerts and thresholds

    Threshold levels are designed to surface problems at two stages:

    1. Warning gives you time to investigate and plan — the server is under pressure but still functional.
    2. Critical means the server is in a degraded state and action is needed to prevent service disruption.

    Health data feeds into the overall server status indicators on the dashboard, so servers with critical metrics are visually prominent.

    Related