Production url is not responding for forestadmin

[Thibault_Manfredi]

Feature(s) impacted

ForestAdmin cannot detect that my backend is up and running for production purposes

Observed behavior

ForestAdmin is stuck the “Set your environment variable” with the error message “it looks like the URL is not responding”

Expected behavior

The URL is set correctly and respond with ssl certificate when I test it on my browser, Forestadmin should be able to detect it, I restart the operation multiple times, replace with good environment variable each time and restart the backend container.

There is the screen where i’m stuck

Context

• Project name: Lexagone
• Team name: Flexper
• Environment name: production

[morganperre]

Hello @Thibault_Manfredi Welcome to our community

• Can you validate that you’re using forest-express-sequelize version 8.4.8 ?
• Do you mind sharing any log from the backend ?
• Can you share try to CURL your server using the APPLICATION_URL to be sure you get a response ?

Kind regards,
Morgan

[morganperre]

So it was a misconfiguration of the APPLICATION_URL on our side. I can’t find any clues on how that could happen.

I notice that the browser received a NS_ERROR_DOM_BAD_URI on the first call to the backend (/forest/authentication).

I think you have a problem comes from your backend container. Can you try to fix it ? (with the information I gave you in MP)

Best regards,
Morgan

[Julien_Martelli]

Hello @morganperre,

Thanks for your answer.

I am now responsible for this task and I don’t have access to the MP you sent to Thibault.

Could you please send it to me too ?

Thanks !

Regards

[morganperre]

Hey @Julien_Martelli,

Here some ref on this topic.

"NS_ERROR_DOM_BAD_URI: Access to restricted URI denied"

[RESOLVED] AJAX: NS_ERROR_DOM_BAD_URI error all of a sudden?

I’m not sure of the root of this issue. Could you check you backend container config ?

You need to be sure that you can call the route {APPLICATION_URL}/forest and receive an HTTP 204 No Content.

Regards,
Morgan

[Julien_Martelli]

Ok thanks, we’ll check this and let you know.

Regards,

[Hsohm]

Hi @morganperre,

I’m in charge for this issue now, I tried to deploy a new forest remote environment with fresh new environment variables but the issue is still the same.

The application url is responding with a 200 http status code and with a “Your application is running” message.

I printed all environment variables using the shell in the kube pod and everything is right configured, I don’t understand why the forest configuration is saying “It looks like the URL is not responding”…

Best regards,
Hugo

[morganperre]

Hello Hugo,

@Hsohm

I’m in charge for this issue now, I tried to deploy a new forest remote environment with fresh new environment variables but the issue is still the same.

Ok, can you give me the name of this new environment ? I will check that the config is ok.

@Hsohm

The application url is responding with a 200 http status code and with a “Your application is running” message.

So you have resolved the NS_ERROR_DOM_BAD_URI error ?

@Hsohm

I printed all environment variables using the shell in the kube pod and everything is right configured, I don’t understand why the forest configuration is saying “It looks like the URL is not responding”…

Can you share the failing call ? I just want to be sure that the right backend is being called.

Best regards,
Morgan

[Hsohm]

The environment is still the same named as “Lexagone”.

I don’t know if I solved the NS_ERROR_DOM_BAD_URI error but I get a 200 on https://admin-dev.lexagone.flexper.xyz

The failing call is displayed in the forest admin process when the env variables for remote environment are available.

Best regards,
Hugo

[morganperre]

Hey again,

@Hsohm

The failing call is displayed in the forest admin process when the env variables for remote environment are available.

You’re stuck on last step of the deployment.

Can you open the network tab of your browser and have a look at the calls please (and share a screenshot here) ?

I just checked and indeed the agent is running on this URL.

Best regards,
Morgan

[Hsohm]

Yes I’m stuck on the last step of the deployment.

I have probably the same network as you with 4 calls and a 200 http status code on the url request.

Best regards,
Hugo

[morganperre]

You should have a call to the agent URL (https://admin-dev.lexagone.flexper.xyz/forest), that not the case ?

Best regards,
Morgan

[Hsohm]

That’s not the case, I don’t have any call to https://admin-dev.lexagone.flexper.xyz/forest, the first one is calling https://admin-dev.lexagone.flexper.xyz.

Best regards,
Hugo

[morganperre]

@Hsohm

That’s not the case, I don’t have any call to https://admin-dev.lexagone.flexper.xyz/forest, the first one is calling https://admin-dev.lexagone.flexper.xyz.

I’m not able to reproduce the same issue. In my case the front is still trying to reach the /forest route.

• Can you try another fresh install ? (You will need to update your FOREST_ENV_SECRET and restart your container)
• Could you open the network tab when you start the process so we could see all calls ?

Let me know if it helps this time.

Plus, I’ll do a feedback to our team because you should be able to relaunch the system that detect the agent is running instead of being stuck at this point.

Kind regards,
Morgan

[Hsohm]

I’ll try to try with a new fresh install and I will list you all the network calls during the process.

Thanks for the feedback to your team to continuously improve the forest admin solution.

Best regards,
Hugo

[Hsohm]

I looked on another forest admin back-end already deployed working on production and it doesn’t call the /forest either.

I don’t think that the problem is about this missing http call.

Here is the backend url for the other project: https://admin-prod.sponsoplus.flexper.xyz.

Best regards,
Hugo

[morganperre]

Hey @Hsohm,

@Hsohm

forest admin back-end already deployed

It only calls the /forest route during initialisation phase (when you first onboard a new project or when you deploy to production/remote).

I try to update your project on my side. Can you try to connect to your Pre-Production environment for Lexagone project ? Let me know what happen (if you can connect to this environment).

Best regards,
Morgan

[Hsohm]

I already tried to connect to our Pre-Production environment for the Lexagone project but I’m still stuck on the last step with this message (I can’t connect to this environment in the project settings):

And yet, the url works well with the same calls as our Forest Admin back-ends that work. https://admin-dev.lexagone.flexper.xyz

[morganperre]

Hey @Hsohm,

A little update on this thread.

After some fast investigation we found some missing allowed-headers in the app.js.

Also, the infrastructure you use has a kubernetes Ingress config that might cause CORS issue.

Lets us know if you found a work around.

Kind regards,
Morgan

[Hsohm]

Hello @morganperre,

The idea during the live debug was the good one, removing our Kubernetes configuration snippet on the nginx ingress worked well to solve the CORS issue.

Thanks and best regards,
Hugo