Error in Lab 11.2

ogbonnaec

Hello Support,
I am encountering error in the lab 11.2.
I had to uninstall linkerd and did 11.1 all over again.
each time I execute the step 11 lab 11.2 command that injects linkerd into the ingress controller immediately my linkerd dashboard that has been working fine from lab 11.1 throws error.
The error does not stay.
I have seen too many people complain about this error but I have not seen any concrete answer on why this is happening or steps to fix this error from the support team hence I started this discussion.
I have been stock here for one week now. Initially I felt i did something wrong but I cleared everything with the help of the linkerd docs and started again.
I would appreciate if this can be fixed

chrispokorni

Hi @ogbonnaec,

I tried to reproduce the issue reported above, however, following all the steps from the latest release of the lab guide no error was displayed. In fact, everything worked as expected.

What is the version of the lab guide (release date)?
What is your Kubernetes version? What is the Linkerd version?
What is the OS of your VMs? What infrastructure hosts your VMs (cloud or local hypervisor)?

What is the output of the following command? Please use the code formatting option when pasting the output in your comment, or upload a text/yaml file with your comment.

kubectl -n linkerd-viz get deploy web -o yaml

If you believe this is a bug, please report the issue on the linkerd project's github page:

https://github.com/linkerd/linkerd2/issues

Regards,
-Chris

ogbonnaec

@chrispokorni,

Thank you for your response. I am using the current version of the lab guide V2024-03-14. I am using a local setup on virtual box. There is nothing wrong with the worker3. It is shutdown

I have also attached the output of this command in the web.txt file attached kubectl -n linkerd-viz get deploy web -o yaml

chrispokorni

Hi @ogbonnaec,

What are the sizes of the VirtualBox VMs (cpu, mem, disk)? How is disk allocated (dynamic or fully allocated)?
How many network interfaces per VM, and what network type they attach to? Is promiscuous mode enabled to allow all ingress traffic?

Regards,
-Chris

ogbonnaec

Hello @chrispokorni ,
My VMs are running with the following resource
CPU: 3
RAM: 8G
Disk: 40
I have two network interfaces one set to bridge and another set to host-only. The bridged interface is the default of kubelet as seen above and promiscuous mode enabled. The second host-only is for my admin purpose so i can ssh and work on my terminal which is more convenient.
Before now everything worked well. I have done the task Lab 11.1 without any issues.
I have also done lab 11.2 until task 11. My ingress is working. It is able route my curl request to the services. I can see the request being served by the pods on the linkerd dashboard.
I noticed a little difference between the image in the lab 11.1 and the deploy/web file of linkerd

Is some entries are missing example is that of the linkerd-metrics-api could this be an issue. The exercise only asked to to edit one line in the deploy/web file.

chrispokorni

Hi @ogbonnaec,

From my previous set of questions:

How is disk allocated (dynamic or fully allocated)?

ogbonnaec

Hi @chrispokorni,
They are dynamically allocated.

chrispokorni

Hi @ogbonnaec,

Kubelet does not favor dynamically allocated disk space, because the control plane only sees the allocated space, not the desired size of the disk. That's where the control plane begins to panic, and throttles the application Pods scheduling process, to the point that some Pods will not get scheduled at all.

Regards,
-Chris