From 1addf536f55761fbb999ab7901e3bacc7eef504e Mon Sep 17 00:00:00 2001 From: SmritiSatya <117705907+SmritiSatya@users.noreply.github.com> Date: Mon, 13 Jan 2025 16:14:07 +0530 Subject: [PATCH 1/2] Content to Linux faults --- .../use-harness-ce/chaos-faults/linux/linux-cpu-stress.md | 7 ++++++- .../chaos-faults/linux/linux-disk-io-stress.md | 6 ++++++ .../chaos-faults/linux/linux-memory-stress.md | 5 +++++ 3 files changed, 17 insertions(+), 1 deletion(-) diff --git a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md index 98d4c7a89f8..f06cbfd647d 100644 --- a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md +++ b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md @@ -37,7 +37,7 @@ This fault uses [`stress-ng`](https://github.com/ColinIanKing/stress-ng), which load Percentage load to be exerted on a single CPU core. - Default: 100 %. + Default: 100 %. 0 refers to no load and 100 refers to full load. workers @@ -81,6 +81,11 @@ spec: The `load` input variable exerts the CPU load (in percentage) per core. +:::tip +- The load generated by the stress process may vary based on factors like overall processor load and the responsiveness of the system scheduler. +- Harness recommends allocating a number of CPUs equal to or greater than the total number of logical CPU cores. +::: + The following YAML snippet illustrates the use of this input variable: [embedmd]:# (./static/manifests/linux-cpu-stress/load.yaml yaml) diff --git a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-disk-io-stress.md b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-disk-io-stress.md index 85681c3958a..08c6fe13c4d 100644 --- a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-disk-io-stress.md +++ b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-disk-io-stress.md @@ -12,6 +12,12 @@ import FaultPermissions from './shared/fault-permissions.md' Linux disk IO stress applies stress on the disk of the target Linux machines over I/O operations for a specific duration. +It aims to consume the I/O bandwidth by performing frequent writes and reads to and from the disk, respectively. Consequently, the size of the file created for this operation changes frequently. + +:::tip +To understand the impact of the fault, check the available I/O bandwidth before and during chaos. +::: + ![Linux disk IO stress](./static/images/linux-disk-io-stress.png) ## Use cases diff --git a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md index 7fff1cf46e0..9d4f3d07035 100644 --- a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md +++ b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md @@ -85,6 +85,11 @@ spec: The `memory` input variable specifies the amount of memory to be filled. +:::tip +- The `stress-ng` package attempts to utilize a percentage of the available free memory using each stressor process. +- Harness recommends setting the number of workers to be equal to or greater than the total number of logical CPU cores. +::: + The following YAML snippet illustrates the use of this input variable: [embedmd]:# (./static/manifests/linux-memory-stress/memory.yaml yaml) From 5f742b43c431ea4cf3b6713f8d37b0a49d7310d2 Mon Sep 17 00:00:00 2001 From: SmritiSatya <117705907+SmritiSatya@users.noreply.github.com> Date: Wed, 15 Jan 2025 17:37:23 +0530 Subject: [PATCH 2/2] Review --- .../use-harness-ce/chaos-faults/linux/linux-cpu-stress.md | 2 +- .../use-harness-ce/chaos-faults/linux/linux-memory-stress.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md index f06cbfd647d..86aebf4d6df 100644 --- a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md +++ b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-cpu-stress.md @@ -83,7 +83,7 @@ The `load` input variable exerts the CPU load (in percentage) per core. :::tip - The load generated by the stress process may vary based on factors like overall processor load and the responsiveness of the system scheduler. -- Harness recommends allocating a number of CPUs equal to or greater than the total number of logical CPU cores. +- It is recommended that the number of CPU cores provided as input should be equal to or greater than the total number of logical CPU cores. ::: The following YAML snippet illustrates the use of this input variable: diff --git a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md index 9d4f3d07035..89e45c2306d 100644 --- a/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md +++ b/docs/chaos-engineering/use-harness-ce/chaos-faults/linux/linux-memory-stress.md @@ -87,7 +87,7 @@ The `memory` input variable specifies the amount of memory to be filled. :::tip - The `stress-ng` package attempts to utilize a percentage of the available free memory using each stressor process. -- Harness recommends setting the number of workers to be equal to or greater than the total number of logical CPU cores. +- It is recommended that the number of workers should be equal to or greater than the total number of logical CPU cores. ::: The following YAML snippet illustrates the use of this input variable: