Unsuccessful Network Delay experiment keeps running after being paused. Deletion is also problematic

[sjafarli]

Bug Report

What version of Kubernetes are you using?
Client version: v1.29.3
Server version: v1.27.10

What version of Chaos Mesh are you using?
GitVersion:"chart-2.6.3", GitCommit:"b159d1c5a361ca350ca7789c9eb7fc917479db39", GoVersion:"go1.19.7"}

What did you do? / Minimal Reproducible Example
When networkchaos experiment is not successful (due to large network delay value which was 5m in my experiment), even if the experiment is paused, it keeps running in the background and events being produced consistently (can be seen from the chaos mesh UI).
On an additional note, deletion is also problematic and takes forever. We remove the finalizer from the resource definition and then try to delete the experiment, it works.

What did you expect to see?
When the experiment is paused, it should not produce any more events.

What did you see instead?
Experiment kept running even if it was paused.

Output of chaosctl

[STRRL]

Hi @sjafarli , could you provide more information to help us address the problem? like:

• how do you measure the network metrics
• the log of chaos-daemon
• chasoctl debug outputs

[sjafarli]

Hi @STRRL , thanks for the reply.

You can quickly set up a test experiment with network delay and give the delay >= 5 m. This will fail due to the limit on max delay value and the experiment will keep producing failed events over and over again and wont be possible to delete the experiment until you delete the pods which the experiment applied on. Then the experiment can also be deleted. I cannot provide the logs since the chaos project i was doing is pretty much done and I figured out how i can basically workaround it