SRE

SRE Server Troubleshooting Guide

A practical troubleshooting workflow for investigating CPU, memory, disk, processes, services, logs and network problems.

Production Troubleshooting Workflow

1. Check CPU

top

Check whether CPU usage is unusually high and identify processes consuming CPU.

2. Check Memory

free -h

Check total, used and available memory on the server.

3. Check Disk

df -h

Check filesystem usage and identify disks approaching capacity.

4. Check Processes

ps aux

List running processes and identify unexpected or resource-heavy processes.

5. Check Services

systemctl status <service>

Verify whether the required system service is running.

6. Check Logs

journalctl -u <service>

Review systemd logs for errors and service failures.

7. Check Network

ss -tulpn

Check listening ports and identify which processes are using them.

8. Check Connectivity

ping <host>

Test basic network connectivity to another host.

Quick SRE Troubleshooting Checklist

# CPU
  top
  
  # Memory
  free -h
  
  # Disk
  df -h
  
  # Processes
  ps aux
  
  # Listening ports
  ss -tulpn
  
  # Service status
  systemctl status <service>
  
  # Service logs
  journalctl -u <service>
  
  # Network connectivity
  ping <host>

DevOpsToolbox Tip

Avoid changing production systems immediately when an alert fires. First collect evidence: metrics, process information, logs, service status and network information. Then identify the likely cause before making a change.