GL325 · 5 days · 8+ hrs hands-on labs

Linux Performance Tuning & Analysis

Available for RHEL· SLES

Master the tools, methodologies, and kernel-level knowledge needed to analyze, monitor, and optimize Linux performance on enterprise systems. Starting with Brendan Gregg's USE Method for systematic bottleneck identification, students build a deep understanding of Linux internals covering kernel architecture, process scheduling, virtual memory, block devices, filesystems, and the I/O subsystem, then apply that knowledge to real-world performance challenges on your choice of RHEL or SLES.

The course covers the full Linux observability stack: process and resource monitoring with ps, top, vmstat, iostat, and mpstat; long-term trending with sar and Performance Co-Pilot; kernel and userspace instrumentation with SystemTap; CPU profiling and flamegraph generation with perf; remote metric gathering with SNMP; and benchmarking with Bonnie++, iperf, and the Phoronix Test Suite. Students also learn to tune key kernel subsystems including CPU scheduling, NUMA memory allocation, huge pages, disk I/O schedulers, filesystem parameters, and network stack configuration using both manual sysctl settings and automated tuned profiles.

Hands-on lab exercises throughout the course let students apply performance analysis and tuning techniques on live RHEL or SLES systems, building the practical expertise needed to diagnose and resolve performance issues in production environments.

Who Should Attend

Linux system administrators, performance engineers, DevOps professionals, and site reliability engineers responsible for optimizing, monitoring, and troubleshooting performance on enterprise Linux servers running RHEL or SLES.

Skills Students Will Gain

Apply the USE Method to systematically identify performance bottlenecks across CPU, memory, network, and storage resources
Monitor CPU, memory, and I/O utilization using vmstat, iostat, mpstat, sar, and top
Instrument the Linux kernel and userspace applications with SystemTap for deep performance analysis
Deploy and configure Performance Co-Pilot for distributed metric collection, inference rules, and historical trending
Profile applications and generate flamegraphs using Linux perf
Configure SNMP agents and gather remote performance metrics across infrastructure
Benchmark system performance using disk, network, and application-level tools including Bonnie++, iperf, and the Phoronix Test Suite
Tune Linux kernel parameters for CPU scheduling, memory management, I/O schedulers, and network performance
Optimize NUMA topology and huge page configurations for memory-intensive workloads
Administer system logging with rsyslog and the systemd journal for performance-related troubleshooting
Trace and debug application behavior using strace, ltrace, and pmap
Apply automated tuning profiles with tuned for common workload patterns

Chapters & Labs

20 labs · 8+ hours hands-on
  1. Tuning Principles & The USE Method 2 topics 2 labs · 65 min
  2. Linux Internals Review 20 topics 2 labs · 40 min
  3. Linux Userspace Architecture 3 topics 3 labs · 35 min
  4. Log File Administration 8 topics 2 labs · 25 min
  5. Monitoring & Troubleshooting 12 topics 4 labs · 85 min
  6. Benchmark Tools 4 topics 2 labs · 90 min
  7. Systemtap 14 topics 1 lab · 30 min
  8. Performance Co-Pilot 9 topics 2 labs · 40 min
  9. Remote Metric Gathering With SNMP 5 topics 1 lab · 20 min
  10. Linux perf 1 topic 1 lab · 45 min
  11. Linux Tunables 10 topics

Apply for partner access to see full topic and lab details.

Prerequisites

Solid experience with Linux system administration including command-line proficiency, process management, storage administration, and networking fundamentals. GL250 (Enterprise Linux Systems Administration) or equivalent experience recommended.