Skip to content

node_disruption.md

Latest commit

 

History

History
17 lines (10 loc) · 1.08 KB

node_disruption.md

File metadata and controls

17 lines (10 loc) · 1.08 KB

Node failure

The nodeFailure field triggers a kernel panic on the node. Because the node will reboot suddenly, the pods running on it (the injection pod included) won't see their status updated for a while. This is why the injection pod can appear as Running or Unknown while it has currently finished the injection. The kernel panic is triggered by using sysrq. The injector uses the following mounts to use it:

  • /proc/sys/kernel/sysrq > /mnt/sysrq
  • /proc/sysrq-trigger > /mnt/sysrq-trigger

⚠️️ Node behavior when using this disruption can differ depending on the cloud provider (node may or may not be replaced, restarted, cordoned, etc.).

Targeting

For clarity purpose, it is mandatory to explicitly set the disruption's level field to either pod or node.

The nodeFailure disruption acts on a node. However, the level field needs to be adapted for the selector:

  • if level: node is set, the selector will target nodes and impact those nodes directly.
  • if level: pod is set, the selector will targets pods and impact the nodes that host them.