Skip to main content

Collect All CPU Profiles From All Cribl Worker Processes on a Linux Worker or Linux Edge Node

  • June 23, 2026
  • 0 replies
  • 1 view

Jessica Bracken

Objective

Collect cpuprofiles from all the Cribl Worker processes on a Linux Worker or Linux Edge node.

Environment

  • Self-hosted Linux Worker node or Linux Edge node

Procedure

  1. SSH into a Linux worker or Linux Edge node. Note: If you are logged in as a different user than the cribl user, run su - cribl
  2. Run this code (from anywhere):
    date +%s; ps -ef | grep -i cribl | grep -v grep | grep -i worker | awk '{print $2}' | while read pid; do echo "Processing: $pid"; /opt/cribl/bin/cribl diag cpuprofile -p $pid; done
    The code will print the execution time in epoch time and list the Worker processes to be run for the cpuprofile. Sample output:
    1776825317 Processing: 2216459 Processing: 2216484 Processing: 2216500

    Be patient and check the diag (ls -latr /opt/cribl/diag/) directory to confirm the cpuprofile is complete.

    Example:

    -rw-r--r-- 1 cribl cribl 352050 Apr 22 02:36 cpu-1776825364383-2216459.cpuprofile -rw-r--r-- 1 cribl cribl 298674 Apr 22 02:36 cpu-1776825372348-2216484.cpuprofile -rw-r--r-- 1 cribl cribl 261213 Apr 22 02:36 cpu-1776825372148-2216500.cpuprofile
  3. Download the profiles from the working directory (for example, use the scp command).
  4. Happy cpuprofile analyzing! You can use the browser’s built-in tools. For example, for the Chrome browser: Chrome tools: (Use CTRL + SHIFT + J on Linux). Go to the performance tab and upload the CPU profile (look for the arrow-up icon on top).

     

Last Validated

Cribl 4.17.0