28 Replies
Not a permanent solution but temporary - if you remove the backend of the URL, it'll load fine again. So like example.com/objects/people is the page it's doing this on, I'd remove /objects/people and hit enter and it loads @ralfdigitalagency
Twenty
A modern open-source CRM
This is the url I'm using now, I then want to change URL (to a specific subdomain I own) but first I wanted to make it work on the localhost url, I'm struggling I tried every guide but not able to find a solution so far
I'm facing with the same issue
Happy to hear I am not the only one, hope someone can help us out
Does a solution exist?
I'm also waiting for the same !
facing this issue currently

Nobody came back to me so far, even in Github, with a solution, I do see severl people are facing the same issue
I don't know if that is a sign of malfuncion, so better move to another tool or it is something can be fixed.
HI @ralfdigitalagency, taking a look at the link you provided:
I can see that your FRONTEND is trying to reach your BE on https://hungry-easley.66-179-210-156.plesk.page//graphql

this domain does not seem to be available
@Harsh23 your issue looks completely different: "User is not logged in". Could you share your server logs ? (depending how you are hosting it, docker logs twenty-server-1 might help you get the logs)
Hello @ralfdigitalagency I don't think your BE is correctly setup https://hungry-easley.66-179-210-156.plesk.page//graphql which result in a 404
https://hungry-easley.66-179-210-156.plesk.page/ is just the default page of plesk your host
Your api served on //graphql does not seem to be up and running
thanks for the reply @charles the domain it is available, if you try to visit it, try it (https://hungry-easley.66-179-210-156.plesk.page/) it is a test subdomain, i wanted to test TwentyCRM there before moving to another permanent subdomain. I have the feeling that the software is correctly installed butthere is something wrong with the pointing. I tried for 3-4 days to solve that with no solution, I also asked a IT tech aquaintance but nothing.
From the guide https://twenty.com/developers/section/self-hosting/docker-compose it is not clear where the application will be accessible from, it mentions http://localhost:3000 but being a VPS should it be http://66.179.210.156:3000/ but the fact is that it is not working in any of these urls: http://66.179.210.156:3000/, http://localhost:3000, https://hungry-easley.66-179-210-156.plesk.page/.
I agree with you @prastoin , how do you suggest I can fix it? Any guide or suggestion to check and potentially fix that?
I don't know anything about plesk, but I would try to first deploy a very simple api to debug my way through your current issue following https://www.plesk.com/blog/product-technology/node-js-plesk-onyx/
Plesk
How to Host Node.js Applications on Plesk Obsidian
On Plesk Obsidian, hosting Node.js applications is easier than ever. See the newest features and installation tips here.
Hey @prastoin,
Thanks for the suggestion! However, the official TwentyCRM guide uses Docker and Docker Compose, so I’m not trying to run Node.js directly on Plesk. The issue might be that Nginx, as a reverse proxy, is not correctly forwarding requests to the Docker containers.
I was thinking of checking the Nginx configuration to make sure it properly routes traffic to the backend. Do you have any advice on this?
Thanks again for your help! 😊
I think that's a good track !
I'm sorry, @prastoin I am a bit confused, as the guide do not mentions Nodejs, why or how should I use that service? Are you suggestin to host the twenty crm without docker, using node.js? Or git extention?
No not at all, I think the issue does not come from twenty image itself ( at least for the moment we can't attest it ) but rather in your deployment process
That's why I was suggesting starting from deploying an hello world node.js ( could be an node.js hello world docker image btw ) to better understands where your issue is along your plesk deployment process
@prastoin I see, thanks, I will look into that, the weird thing is that many people are facing my very same issue.
This is the setting of the Docker Container Server of Twenty CRM, are these settings correct?


I will look into that, the weird thing is that many people are facing my very same issue.If you're talking about people in this thread, I think they mainly came to the conclusion they have the issue as you because of your first message and help thread title which are very generic But in my opinion have very high chances to be different root causes It seems like several env vars value are missing in your configuration If I was you I would try reaching to my docker image logs and see if some are required and missing making the twenty-server bootstrap fail But again I would follow above reco in the first place
I finally got mine running. Try removing the trailing / in your SERVER_URL
Great! @craigzyc what do you put into the SERVER_URL field exactly?
I'm did not use Docker while setting up and hence using homebrew, postgres and redis. But the errors I faced while setting up postgres server -
- query failed: CREATE EXTENSION IF NOT EXISTS "vector"
error: error: extension "vector" is not available
Failed to perform 'create extension "vector (pgvector)"':
-ERROR: schema "core" does not exist
-Error executing migration QueryFailedError: relation "core._typeorm_generated_columns_and_materialized_views" does not exist
and then USER is not logged in.
Hello @Harsh23 please open a dedicated thread in #❓︱help with your current issue and related logs information etc
Mine is running on my local network on a docker container for testing so it was just http://192.168.0.2:3000
I originally accidently had localhost in there but it is not on my local machine so that obviously didnt work.
After changing it to the IP I noticed the request had a double // and failed. Removing the trailing slash made it all work
Had a double slash here

The only other major issue I had was with the postgres database I had to manually change the password.
psql -U postgres
ALTER USER postgres WITH PASSWORD 'newpassword';