Kodekloud Challenge #1 issue with creating the pod - issue with command

[mmkmou]

Hi @boy
As the PV is already claimed you have two options :

• Remove it and recreate it as it was
• try to unclaim it by patching the PV see the following discussion
  Need to know how I can 'unclaim' a PVC once it's been attached to a PV

Regard

[Carim-Manzur]

Hello! It would be great to “understand” the problem.

I think the real problem is that the /site directory needs to be cleaned, so the solution will be:
1- Delete the full directory on the k8s node with rm -rf /site
2- Delete the pod, pvc and pv
3- Create pv, pvc and pod

[George-Miscalencu]

This is so damn frustrating.

[hpeters]

@klein.joshua.a

apiVersion: v1
kind: Pod
metadata:
  creationTimestamp: null
  labels:
    run: jekyll
  name: jekyll
  namespace: development
spec:
  initContainers:
  - name: copy-jekyll-site
    image: kodekloud/jekyll
    command: 
     -  jekyll
     -  new
     -  /site
    volumeMounts:
     - mountPath: "/site"
       name: site
  containers:
  - image: kodekloud/jekyll-serve
    name: jekyll
    resources: {}
    volumeMounts:
     - mountPath: "/site"
       name: site
  dnsPolicy: ClusterFirst
  restartPolicy: Always
  volumes:
   - name: site
     persistentVolumeClaim:
       claimName: jekyll-site
status: {}

I have the same issue, with the above yaml

[mmkmou]

Hi @hpeters,

How did you configure your PVC?

[DVS_Dee]

Same issue here as described by OP.

[raghumanne]

First see if the pv and pvc are properly up and bounded.
I observed even though the command: [ "jekyll", "new", "/site" ] is fine, the check won’t pass until the command is executed.

[jdomm95]

I had the same issue as above. I waited a good 20 minutes and the pod never finished. As recommended, I reset the lab and tried the git source and it worked! But my yaml’s were the same as the git ones. I am wondering if the issue is caused if you create a pod with something incorrect, then delete the pod, fix the problem and create the pod again. That is the only thing different in me doing it myself and when I did it with the git source after resetting the lab. Just a suggestion for kk to look into to see why so many people are getting this issue, even after waiting 2 minutes for the pod to start up.

[tsanghan]

@jdomm95 ,

There is NO issue with the lab.

[Alistair_KodeKloud]

Occasionally this pod does fail to come up. Nobody has determined exactly why yet. Normally reset the lab and try again (copy the manifest you created to a notepad first to save time)

[tsanghan]

@Alistair_KodeKloud ,

The Pod does take sometime to come up, but NEVER failed to come up in my test run of 30 times of doing the challenge.

All test run have Pod in Ready state within 30s of kubectl -n development wait --for=condition=Ready pod/jekyll, except Test run #29.

I will attribute the issues reported in this thread/post above of the case of,

1. Incorrect PVC
2. Incorrect namespace
3. and any other human errors (which no body acknowledge they had because they are always right)

Rather then attributing to Ghost in the Shell.

A curious case of Occam's Razor?

[Alistair_KodeKloud]

I see you’re running my automation script. TBF I haven’t seen it fail with that either.

[tsanghan]

No, I am not using your automation script. I ran my own script.

[Alistair_KodeKloud]

At least that verifies the process from two angles.

[mihai.cazacu]

Hi!
This problem still exists:

[mihai.cazacu]

[mihai.cazacu]

The issue can be reproduced by resetting the lab and running the following commands:

[Alistair_KodeKloud]

[Alistair_KodeKloud]

I suspect, but I cannot prove that there may be an intermittent network issue between the cluster and the internet, so ruby fails to download gems.
Reset the lab over and over and eventually it should work

[carolinerj]

you are totally right !