Droplet is blank
Hi all I recently tried to deploy twenty to digital ocean on a droplet however it is not working after I'm run everything. When I go to the address in my web browser the page is blank white. Any help will be appreciated
27 Replies
Hey, could you post some logs from your server and client as well?
I've had the same issue as well.
after following the one command instructions on the following page:
https://docs.twenty.com/start/self-hosting/docker-compose
Is there a step that's missing?
Welcome | Twenty - Documentation
Twenty is a CRM designed to fit your unique business needs.
I am also experiencing this issue. I followed the troubleshooting steps but no luck π¦
Your frontend is likely not able to connect to your backend. It's usually an issue with DNS or SSL (https). Could you show your developer console ?
(we are not covering DNS and SSL setup in our self-hosting doc at the moment as this is very vendor dependant)
But happy to give a few hints if I can
I was suspecting sth like that is happening. Will send a screenshot once I get back home.
Iβm having the same issue. @charles based on what you said if I understand I need SSL and a DNS record to show the CRM? If Iβm just trying to access from my IP it will just show the white screen?
Could you post a screenshot from your developer console? It's likely to be DNS/SSL issue but I can give you a more accurate response with a bit more info from your dev console
screenshot from web console:
showing some CORS errors
that's the network tab π Could you show me the console?
yes, sorry π
here it is:
So your FE is trying to reach your server on localhost:3000 which is not right
make sure that your REACT_APP_SERVER_BASE_URL is properly setup π
you should not have any DNS issue as you are using IP
you are not using SSL which is a big red flag but it should not prevent twenty from working π
thanks for the hints, I'll tinker a bit now and see if I can manage, thx for being so responsive btw
REACT_APP_SERVER_BASE_URL what shoud this env be? where can I find the setup
Same here I would say (assuming you want to self-host, not to contribute)
Still haven't managed to figure it out, is there something I am missing from the setup? I only have SERVER_URL in the env, not REACT_APP_SERVER_BASE_URL
changed the SERVER_URL to my Droplet's IP also :3000
Now I am able to see the page, albeit no ssl, and now next issue is with the sign up, I tried adding smpt for resend but am getting QueryFailedError.
Hello, I managed to set up everything.
I followed these steps:
docker exec -it twenty-server-1 yarn
docker exec -it twenty-server-1 npx nx database:reset
but the second command I had to do:
docker exec -it twenty-server-1 yarn database:init:prod
and also had to upgrade to 8GB RAM during the execution of the command. But now everything works accordingly.
glad it works π on production, you should not have to do yarn install and to use npx
npx is a dev tool that caches stuff because dev environment is not optimized
production container come with pre-built code
in production container, you should use: yarn command:prod, yarn database:init prod
it is also less resources heavy
HI I am little confused with what you said. So I start the two docker containers however that database is not initialized properly so I run docker exec -it twenty-server-1 yarn database:init:prod
but I get this as a response:
<--- Last few GCs --->
0.[1447:0x764b82a56020] 36268 ms: Mark-sweep (reduce) 983.3 (1001.2) -> 982.8 (1001.5) MB, 933.5 / 0.0 ms (+ 58.0 ms in 14 steps since start of marking, biggest step 5.9 ms, walltime since start of marking 1062 ms) (average mu = 0.478, current mu = 0.31[1447:0x764b82a56020] 38305 ms: Mark-sweep (reduce) 984.0 (1001.7) -> 983.3 (1002.0) MB, 1496.0 / 0.0 ms (+ 69.6 ms in 15 steps since start of marking, biggest step 5.8 ms, walltime since start of marking 1634 ms) (average mu = 0.348, current mu = 0.2
<--- JS stacktrace --->
1GB is too small, you should have at least 2GB for the server to properly run
I am using 2gb RAM droplet
your node seems to be caped to 1000Mb according to your logs
you can play with the limit using: "NODE_OPTIONS": "--max_old_space_size=5000"
(env variable)
Let me try and change it to 2gb and then run again
Ok so I changed the variable in docker compose enviroment variable and then ran it and this was the error:
Is there a way to innit the database by directly interacting with the database container
Your database is not properly initialized
you should have a pre-existing core and metadata schema, it's likely that the initilization as failed when starting the postgres container (usually it happens when changes are made in the env variables of this container)
could you share you docker-compose?
It seems that you need a 4gb server to run the CRM
At least to initialise the database> I have managed to get it up now
That's heavy, we should look into that