Possible wrong assumption in LFD259 lab handbook

matthyx

Hello,

I am currently training for the CKAD, and have read the following in the lab handbook:

Where the readinessProbe is checked prior to being ready, then not again, the livenessProbe continues to be checked.

This is wrong as far as I know... normally both probes are always checked, but the action taken when they fail is different.
Could you reach the writer and make sure this is correct?

Thanks,
Matthias

FaridSaad

Experienced the same. readinessProbe status changed back from pass to fail later through the container's lifetime during my tests.

chrispokorni

Hi @FaridSaad,

Throughout Lab Exercise 3.3 the readinessProbes are expected to fail every time a Deployment is created. More precisely, the readinessProbes of the try1 deployment (from simpleapp.yaml) are supposed to fail at steps 2,3,4 and pass at steps 5,6,7,8. Then again readinessProbes fail at steps 10,11 (as the deployment has been re-created) and pass at steps 12,13.

Did you experience other probe failures?

Regards,
-Chris

gerardocl

I also think that the statement in the LFD259 slides is misleading since the K8s docs state the contrary:

Note: Readiness probes runs on the container during its whole lifecycle.

https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/#define-readiness-probes

What I get from the K8s docs is that the readinessProbesare checked again every periodSeconds (default value of 10); in the same fashion as livenessProbes.

chrispokorni

Hi @gerardocl,

You are correct, the readinessProbe is checked and re-checked based on the value of the periodSeconds property, according to the documentation and the readinessProbe box from the course.

Regards,
-Chris

gerardocl

Great! Will the statement be updated in the next revision of the course material? It is still there in Chapter 3.

serewicz

Yes.