Why Is My Server Load High and What Should I Do?

Server load reflects how much work (CPU, memory, disk I/O, and processes waiting for resources) your server is handling at a given moment. On Linux, the load average (1/5/15-minute figures) is a quick health signal — as a rule of thumb, a sustained load average above your CPU core count means requests are queueing.

Common causes

  • Traffic spikes (legitimate or bot/scraper traffic)
  • Inefficient database queries (missing indexes, N+1 query patterns)
  • Memory leaks in a long-running process
  • A runaway cron job or background worker
  • Undersized hosting for current traffic

Diagnosing it

  1. Check top/htop to see which process is consuming resources
  2. Check slow query logs on your database
  3. Review recent deploys — a new feature or query is a common trigger
  4. Check for bot traffic in access logs

Monitoring it proactively

Set alert thresholds on load average, memory usage, and disk space so you find out before users experience slow page loads or timeouts, not after.

FAQ

Is high load the same as downtime?

No — a server can be fully "up" while responding slowly under load. Response-time monitoring catches this even when uptime checks still pass.

Updated at: .