ustomDomainCreate fails for all domains — workspace-level authorization block (trace IDs included)
uparrowinc
HOBBYOP

a month ago

Hi Railway Support,

I upgraded to Pro today and the issue persists. I need an engineer to check

backend logs for a workspace-level authorization block preventing all custom

domain creation.

WHAT IS HAPPENING

-----------------

The CustomDomainCreate mutation fails immediately with "Problem processing

request" for every domain I attempt to add — including completely random,

unregistered domains (e.g., random123456789.example.com). The failure occurs

before any DNS or TXT verification step.

The customDomainAvailable query also returns "Not Authorized"

(INTERNAL_SERVER_ERROR) for my workspace, confirming the block is at the

authorization/pre-validation layer, not at the DNS layer.

WHAT I HAVE RULED OUT

---------------------

- DNS: CNAME is globally propagated and confirmed correct.

- Domain-specific lock: Random unregistered domains fail identically.

- Environment corruption: Brand new projects/environments fail the same way.

- Quota: I have 0 custom domains attached across all projects (limit is 20 on Pro).

- Previous provider: Replit has fully purged the domain from their edge.

WORKSPACE AND PROJECT DETAILS

------------------------------

Workspace ID: 0beef95e-b73c-458c-b739-a16a5136dfe8

Project ID: 8407db4c-8d56-4b0b-8e18-adcee20ef05d

Service ID: 9ccf3e56-bc0c-4e88-b084-c53e42a4ae5b

Environment ID: 1139ceb5-d1c2-42fe-a433-2413242c66b0

Domain: ai.uparrowinc.com

TRACE IDs FOR BACKEND LOG LOOKUP

----------------------------------

7893185092485050111 — CustomDomainCreate, post-Pro upgrade (just now)

8582439880873192796 — CustomDomainCreate, pre-upgrade, target domain

2461908427945489241 — CustomDomainCreate, pre-upgrade, random domain

2482298912339763953 — customDomainAvailable query, "Not Authorized"

Please check these trace IDs in your backend logs and lift whatever

authorization or workspace-level flag is blocking domain creation.

Thank you.

Solved$20 Bounty

Pinned Solution

In your input, you need to provide the following:

  • environmentId
  • projectId
  • domain
  • serviceId
  • targetPort

7 Replies

Railway
BOT

a month ago

This thread has been marked as public for community involvement, as it does not contain any sensitive or personal information. Any further activity in this thread will be visible to everyone.

Status changed to Open Railway 26 days ago


Status changed to Solved uparrowinc 26 days ago


uparrowinc
HOBBYOP

a month ago

This is not for public view this was for an engineer and no bounty - terrible support upgrading to PRO


Status changed to Awaiting Railway Response Railway 26 days ago


uparrowinc

This is not for public view this was for an engineer and no bounty - terrible support upgrading to PRO

uparrowinc
HOBBYOP

a month ago

I tried resubmitting this please fix it


uparrowinc
HOBBYOP

a month ago

I went back to Replit for time being and want to come back to Railway once issues is solved


Can you share your GraphQL query?


Status changed to Open 0x5b62656e5d 26 days ago


Make sure your token isn't scoped to another project/workspace as well.


uparrowinc
HOBBYOP

a month ago

Hi,

0x5b62656e5d

MODERATOR

The GraphQL mutation used was:

mutation CustomDomainCreate($input: CustomDomainCreateInput!) {

customDomainCreate(input: $input) { id domain }

}

Variables:

{

"input": {

"domain": "ai.uparrowinc.com",

"environmentId": "1139ceb5-d1c2-42fe-a433-2413242c66b0",

"serviceId": "9ccf3e56-bc0c-4e88-b084-c53e42a4ae5b"

}

}

The token is a personal access token from Account → Tokens. It is not

project-scoped. I confirmed this by successfully querying my account

profile, workspace, and deployment history with the same token.

To rule out token scope as a cause: I also tested with a completely

random domain (random123456789.example.com) against a brand new project

and environment created minutes prior — same "Problem processing request"

error. A scoped token would return an authorization error referencing the

project, not a generic processing failure across all projects including

new ones.

The customDomainAvailable query also returns "Not Authorized"

(INTERNAL_SERVER_ERROR) for my workspace regardless of domain.

Please look up traceId 2482298912339763953 in your backend logs —

that is the customDomainAvailable failure and should show exactly

where the authorization is being rejected.

Thank you

Attachments


In your input, you need to provide the following:

  • environmentId
  • projectId
  • domain
  • serviceId
  • targetPort

Status changed to Solved 0x5b62656e5d 4 days ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...