lab 3.2 post simpleapp deployment not working

santoshsub123

wherever simpleapp is referenced in future labs i am getting error imagebackoff in pod status and when i describe i am getting error

Warning Failed 14s (x2 over 31s) kubelet, 2b7fb2b29d1c.mylabserver.com Failed to pull image "10.100.28.200:5000/simpleapp": rpc error: code = Unknown desc = Error response from daemon: manifest for 10.100.28.200:5000/simpleapp:latest not found: manifest unknown: manifest unknown

Normal BackOff 0s (x3 over 30s) kubelet, 2b7fb2b29d1c.mylabserver.com Back-off pulling image "10.100.28.200:5000/simpleapp"
Warning Failed 0s (x3 over 30s) kubelet, 2b7fb2b29d1c.mylabserver.com Error: ImagePullBackOff

kindly suggest next steps how to debug this issue.

Thanks

serewicz

Hello,

There are a few reasons this could happen. The most common one I have encountered is when docker is configured on the master but not on the worker. Second most common is a typo in the image name.

When you run sudo docker image do you see the simpleapp image? If you use sudo docker pull 10.100.28.200:5000/simpleapp on the master and worker does it work?

Regards,

suser

Hello,
I keep having the same problem after rebooting the master node. Apparently the image needed for try1 deployment do not persists on reboot.

[email protected]:~$ kubectl get po
NAME READY STATUS RESTARTS AGE
nginx-b68dd9f75-vq6vr 1/1 Running 1 21h
registry-cbc9b4779-5qqq5 1/1 Running 5 2d23h
try1-85f7fc6ddf-8m766 0/1 ImagePullBackOff 0 20h
try1-85f7fc6ddf-9lbc7 0/1 ImagePullBackOff 0 20h
try1-85f7fc6ddf-c958k 0/1 ImagePullBackOff 0 20h
try1-85f7fc6ddf-fkxg5 0/1 ImagePullBackOff 0 20h
try1-85f7fc6ddf-mkmgs 0/1 ImagePullBackOff 0 20h
try1-85f7fc6ddf-x5snl 0/1 ErrImagePull 0 20h

The repository looks good and running.
What I have to do is to push again the needed image:
sudo docker push 10.107.241.131:5000/simpleapp

After this the try1- pods start one after another in few seconds.

What to do to fix the image persistence?

Stefan

chrispokorni

Hi @suser,

The registry is cleaned up at reboot because the Persistent Volume uses /tmp/... as the hostPath: . You can edit your vol1.yaml and vol2.yaml manifests and update the path: property found under hostPath: to point to a different host directory.

Regards,
-Chris

santoshsub123

Hi @chrispokorni to fix persistence do we need to create that directory on container using exec please advise.
thanks santosh

chrispokorni

Hi @santoshsub123,

Creating the hostPath directory anywhere on the Node/host is fine, as long as it is not in /tmp.

Regards,
-Chris