Closed umarfchy closed 2 months ago
matbmoser
commented
3 months ago Hi @umarfchy,
You need a keycloak instance running to use the front end correctly.
The realm are in our deployment folder.
It is also depending on the data space central services to work. Which can be found in the tractus-x umbrella.
I can't help you much right now since, if the configuration is not filled correctly,
Tell me if it works!
saudkhan116
commented
3 months ago Hi @umarfchy,
The frontend is protected from authentication mechanism from Keycloak, and the system depends on central services.
However, you could still run the frontend component (without helm and docker).
If you still need any support, please let me know!
umarfchy
commented
3 months ago Hi @saudkhan116, I've followed your instructions and tried spinning up the server locally. However, faced another error,
To be exact, I've started a keycloak server using the following command docker run -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=pass -p 8088:8080 quay.io/keycloak/keycloak:25.0.2 start-dev. Then moved to dpp-frontend directory, installed keycloak-js, installed dependencies, and thereafter served the application using npm run serve.
Also, I was wondering if there is any helm chart that can be used to spin up the entire infra locally that has all the components in this diagram. Ideally, I am looking to deploy this setup in environments like Minikube. My goal is to gain a better understanding of how the overall architecture functions in practice.
saudkhan116
commented
3 months ago It seems like an issue with the authentication and backend componet. The backend component cannot work without the dataspace components (shared services) and EDCs.
There is currently no helm chart available for the complete DPP infra setup in this repo, except the umbrella helm chart that spins up the dataspace shared service for you, and then you can configure DPP and make a connection.
I know, its a bit complex to understand, but currently is the only way.
Thats actually a good point for us to think about how DPP can be setup as a standalone and run locally via helm deployment.
umarfchy
commented
3 months ago Hi @saudkhan116, I took some time to play with the umbrella project. Using this tutorial I've deployed the stack in minikube with the following helm command
helm install --set centralidp.enabled=true,managed-identity-wallet.enabled=true,dataconsumerOne.enabled=true,tx-data-provider.enabled=true umbrella tractusx-dev/umbrella --namespace umbrella --create-namespace
I tried multiple times and it times out every time. However, upon checking the pods, I see all pods are running.
Assuming all pods are deployed, I tried accessing http://centralidp.tx.test/ which gave 404 Not Found.
Assuming there was an issue with my DNS and Ingress, I tried directly forwarding the service and pod (k port-forward pods/umbrella-centralidp-0 8080:8080), to see if I could get anything in the browser. However, in both cases, it showed resources not found.
Is there something wrong with my process? I assume this is the first step in preparing the data space. After deploying the dataspace, I need to deploy the edc provider and edc consumer. Once all of these are deployed, I can use the URLs/IPs to connect the dpp backend. Let me know if I'm moving in the right direction.
I appreciate the help.
saudkhan116
commented
3 months ago Hi @umarfchy,
If you enable centralidp, you also need to enable other components portal (backend + frontend), sharedidp, managed-identity-wallet as these componets depends on one another. May be centralidp doesn't find components to access and the condition is timed out, not sure about the actual issue. Could you please enable them from the helm values files and run in a debug mode e.g., helm install --debug umbrella -f values.yaml .
The centralidp is usually accessible at localhost:8080/auth/ in browser if you enable port fowarding k port-forward pods/umbrella-centralidp-0 8080:8080. In oder to connect these services, first they must be accessible inside the cluster, I would suggest to create a debug pod in a cluster and check their accessibility. If everything works, then configure your DNS and ingress settings to make these services accessible outside the cluster.
The edc consumer and provider already comes with the dataspace. You dont need to spin them up separately.
Hope this helps you!
umarfchy
commented
3 months ago Hi @saudkhan116 , thanks for the response. I installed the helm chart enabling the following parameters:-
centralidp.enabled=true
managed-identity-wallet.enabled=true
dataconsumerOne.enabled=true
tx-data-provider.enabled=true
portal.enabled=true
sharedidp.enabled=truehelm install --set centralidp.enabled=true,managed-identity-wallet.enabled=true,dataconsumerOne.enabled=true,tx-data-provider.enabled=true,portal.enabled=true,sharedidp.enabled=true umbrella tractusx-dev/umbrella --namespace umbrella --create-namespace --debug The installation was not complete and there were pods created as before. You can also see the debug log here.
I tried forwarding port using k port-forward pods/umbrella-centralidp-0 8080:8080 however, it showed Resource not found.
Also, I noticed that one of the post install jobs was failing. The log seems to indicate issue with connectivity.
| URI: | /management/v2/catalog/request |
|---|---|
| STATUS: | 500 |
| MESSAGE: | Internal Server Error |
| SERVLET: | EDC-management |
Also, I created a nginx pod and tried connecting to central idp service from within the pod. However, this also showed resource not found.
root@nginx:/# curl 10.244.0.65:8080
<html><body><h1>Resource not found</h1></body></html>
I haven't changed any default values apart from enabling the parameters. Wondering if there are additional changes that are required 😅
saudkhan116
commented
3 months ago Hi @umarfchy from the logs, it seems like a connectivity issue. Please have a look into the deployment docs here: https://eclipse-tractusx.github.io/docs/tutorials/e2e/connect/deployComponents
saudkhan116
commented
2 months ago Since, the required concerns are considered, closing this issue.
Current Behavior
I want to test out the DPP on minikube. I created a pv and installed the app application using the command
helm install digital-product-pass tractusx/digital-product-pass. The applications are running the application and forwarding the port for the DPP frontend, I'm getting the following error in the browser console.Error:
Failed to load module script: Expected a JavaScript module script but the server responded with a MIME type of "text/html". Strict MIME type checking is enforced for module scripts per HTML spec
Note that JavaScript is enabled and tested in both normal and incognito mode.
Expected Behavior
I was expecting to run the frontend on the forwarded port locally.
Steps To Reproduce
System Info: minikube version: v1.33.1 Ubuntu 22.04.4 LTS Operating System Chrome & Firefox Browser