Could you update your document about ingress-nginx install

[wss29]

hi,
Your document about installing ingress-nginx is a little out of date. Simply setting allow-snippet-annotations to true is no longer sufficient to enable the nginx.ingress.kubernetes.io/configuration-snippet annotation. I struggled for several days trying to understand why allow-snippet-annotations: "true" didn't work when installing my-theia-cloud component, until I finally found the answer on ingress-nginx issure

using follow commands can avoid configuration-snippet annotation error when install my-theia-cloud

helm upgrade --install ingress-nginx ingress-nginx \
 --repo https://kubernetes.github.io/ingress-nginx \
 --namespace ingress-nginx --create-namespace \
 --set controller.config.allow-snippet-annotations="true" \
 --set controller.config.annotations-risk-level="Critical"

This can be a headache for new users, especially when they follow the ingress-nginx Helm installation guide directly and then proceed to install my-theia-cloud. so if possible, could you update your document about the ingress-nginx install?

Aside from the previous issue, only one problem is left when installing the my-theia-cloud component. As shown in the image

I'm encountering the error: path /trynow(/|$)(.*) cannot be used with pathType Prefix. Is this related to the Kubernetes or ingress-nginx version? The issue originates from the landing-page-ingress-path-based.yaml file, but I don't know how to resolve it. I've already opened a discussion on GitHub #392. This is the final step to getting Theia-Cloud working, so any help would be greatly appreciated."
my current Kubernetes version is 1.31.0,

[jfaltermeier]

Hi, Thanks for reporting this. We haven’t updated to the latest 1.12 NGINX ingress controller in our demo/tests setups yet. The safest versions for Theia Cloud compatibility can be found in our Terraform configurations: https://github.com/eclipse-theia/theia-cloud/blob/main/terraform/modules/helm/main.tf#L107.
We typically update versions around the time of a new release, so this new NGINX ingress release is too recent for our current setup.

It seems the 1.12 release of the nginx ingress controller introduces some breaking changes, including the issue you reported and this one here: kubernetes/ingress-nginx#12637, which appears similar to your report.
This can be disabled for now: https://kubernetes.github.io/ingress-nginx/user-guide/nginx-configuration/configmap/#strict-validate-path-type

We will update our guides soon

[wss29]

hi @jfaltermeier
I upgraded to version 4.11.3 and the installation was successful. but it still has a warning about path prefix

However, I still cannot access the TryNow website. I have checked the status of all installed components, and everything seems to be running correctly. Did I miss any setup steps, or is there a version mismatch? This issue occurs when I try to access https://192.168.0.140.sslip.io/trynow/.

I noticed a container named image-preload-0, which uses the image docker.io/theiacloud/theia-cloud-demo. This image appears to be the theia-cloud-demo for the TryNow web demo. However, the status of image-preload-0 is "exited." Why is this the case? Could it be related to the pathType: Prefix configuration, or does it depend on a specific version of something?

[lucas-koehler]

Hi @wss29 the pre-loading container is only used to force loading the application image on the nodes. With this, sessions start faster. This is independent of any actual running session pods, the landing page or other services.

I noticed that your values and your address to access the landing page do not match.
You stated you got a 404 for https://192.168.0.140.sslip.io/trynow/.
However, in your values, you use 192.168.1.140.sslip.io as the baseHost. Could this be the problem?

[wss29]

hi @lucas-koehler
I follow the install and setup setps


In the values.yaml file, the basehost is set to 192.168.1.140.sslip.io, but access should be directed to https://192.168.0.140.sslip.io/trynow/. Is this not correct?

[lucas-koehler]

Hi @wss29 ,

In the values.yaml file, the basehost is set to 192.168.1.140.sslip.io, but access should be directed to https://192.168.0.140.sslip.io/trynow/. Is this not correct?

No, this is not correct because the IPs do not match. In the basehost you use 192.168.1.140.sslip.io but then try to access https://192.168.0.140.sslip.io/trynow/ (Notice the difference in the third segment of the IP address. Once you have 1 and once you have 0). However, they need to match because the NGinX Ingress will only react to the host configured in the values.

[wss29]

Hi @lucas-koehler
I notice the problem, I will try to check my values.yaml the day I go to work.thanks very much

[wss29]

hi @lucas-koehler
After changing the IP address to 192.168.0.140 in the settings, I was able to access the web. Thank you for pointing out my oversight regarding the details.