For information on how to use perf, refer to the
perf manual
and the kernel
sysctl settings.
Enable kernel crash dump collection
In the rare event of a kernel panic, enabling kernel crash dump collection can
help you and Cloud Customer Care determine the root cause. To enable crash dumps,
see
User Guide: Kernel Crash Dump Collection for COS
in the open source documentation. This user guide explains how to collect kernel
crash dumps on GKE nodes with Container-Optimized OS or
Container-Optimized OS instances created from Compute Engine.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-08-25 UTC."],[[["\u003cp\u003eThe \u003ccode\u003eperf\u003c/code\u003e tool is pre-installed on Container-Optimized OS.\u003c/p\u003e\n"],["\u003cp\u003eYou can directly execute \u003ccode\u003eperf\u003c/code\u003e commands on your \u003ccode\u003ecos\u003c/code\u003e instance via the command line.\u003c/p\u003e\n"],["\u003cp\u003eElevated permissions might be required to enable performance events, which can be achieved by using the provided command.\u003c/p\u003e\n"],["\u003cp\u003eFor guidance on utilizing \u003ccode\u003eperf\u003c/code\u003e, consult the \u003ccode\u003eperf\u003c/code\u003e manual and kernel sysctl settings.\u003c/p\u003e\n"]]],[],null,["# Use debugging tools\n\nThis page explains tools that you can use to debug\nContainer-Optimized OS.\n\nRun the `perf` tool on Container-Optimized OS\n---------------------------------------------\n\nBy default, [`perf`](https://perf.wiki.kernel.org/index.php/Main_Page)\nis installed on Container-Optimized OS. You can run `perf` on your `cos`\ninstance from your shell command line.\n\nIf you have limited privilege, you might need to enable the use of the\nperformance events system using the following command: \n\n echo -1 | sudo tee /proc/sys/kernel/perf_event_paranoid\n\nFor information on how to use perf, refer to the\n[`perf` manual](http://man7.org/linux/man-pages/man1/perf.1.html)\nand the kernel\n[sysctl](https://www.kernel.org/doc/Documentation/sysctl/kernel.txt) settings.\n\nEnable kernel crash dump collection\n-----------------------------------\n\nIn the rare event of a kernel panic, enabling kernel crash dump collection can\nhelp you and Cloud Customer Care determine the root cause. To enable crash dumps,\nsee\n[User Guide: Kernel Crash Dump Collection for COS](https://cos.googlesource.com/cos/overlays/board-overlays/+/refs/heads/master/project-lakitu/docs/kdump.md)\nin the open source documentation. This user guide explains how to collect kernel\ncrash dumps on GKE nodes with Container-Optimized OS or\nContainer-Optimized OS instances created from Compute Engine."]]
