Amazon Linux Scripting, Automation & Performance Monitoring

Table of Contents

1. Challenge

   Create and schedule automated service health-check scripts using Bash, cron, and systemd timers

   • Write a Bash script that monitors httpd, mariadb, and a custom service, logs structured status lines, and publishes an SNS alert when any service is inactive.
   • Schedule the script with a cron job in /etc/cron.d/ and verify execution through the system journal.
   • Refactor the schedule as a systemd timer with randomized delay and unified journal logging, then compare the two scheduling mechanisms.
2. Challenge

   Register and manage a fleet of EC2 instances using AWS Systems Manager Run Command

   • Confirm that all three fleet instances are registered with Systems Manager and reporting as Online.
   • Send a parallel Run Command to apply security patches and restart httpd across the fleet, then review per-target status and output.
   • Diagnose the failure on fleet-node-2, identify the broken httpd systemd drop-in override as the root cause, apply a targeted fix with a follow-up Run Command, and rerun the patch to bring the fleet to a consistent state.
3. Challenge

   Diagnose performance bottlenecks and configure CloudWatch Agent for custom OS-level metric monitoring

   • Reproduce a CPU spike with lab-stress.sh and trace it to the responsible processes using top, vmstat, and ps.
   • Reproduce an I/O spike and identify the runaway disk activity with iostat.
   • Apply persistent sysctl network buffer tuning and validate it against an iperf3 server on fleet-node-1 before and after.
   • Configure the CloudWatch Agent to publish mem_used_percent and disk_used_percent under a custom namespace, then create an alarm that transitions from INSUFFICIENT_DATA to OK once metrics flow.