2 months ago
I've read all the other issues regarding this and am still unable to solve ECONNREFUSED after deployment.
Things are working in my local enviro with docker REDIS + my worker.
Using:
- Express server
- BullMQ worker
- Redis
I've tried:
1. Setting family: 0
2. Adding a 3 second delay before getting the REDIS_URL
env var and starting the BullMQ worker
Both my BullMQ worker and Redis DB are deployed to Railway, so I am trying to use the REDIS_URL
instead of the REDIS_PUBLIC_URL
Error: connect ECONNREFUSED fd12:4cbb:1509::e7:a1ad:176d:6379
at TCPConnectWrap.afterConnect [as oncomplete] (node:net:1610:16) {
errno: -111,
code: 'ECONNREFUSED',
syscall: 'connect',
address: 'fd12:4cbb:1509::e7:a1ad:176d',
port: 6379
}
I've tried to solve this for hours, but must not be understanding some aspect of it. Please help ^^
FYI: I've removed the deployment to prevent it from spamming the error for hours on end.
5 Replies
2 months ago
Apologies but this looks like an issue with the application level code. Due to volume, we can only answer platform level issues here
I've made this thread public so that the community might be able to help
Status changed to Awaiting User Response railway[bot] • 2 months ago
jake
Apologies but this looks like an issue with the application level code. Due to volume, we can only answer platform level issues hereI've made this thread public so that the community might be able to help
2 months ago
Isn’t everything posted here “application level code”. It’s code to connect to your platform.
We’re looking to upgrade our tier in the next month or so, but not if we’re unable to get this working.
Documentation needs to be clearer about connecting to Redis via private network as many people are having issues.
Status changed to Awaiting Railway Response railway[bot] • 2 months ago
2 months ago
Can I get another EMPLOYEE to help me with this? There are many similar threads which seem to have been assisted by employees but my post gets a non answer.
What do I have to do to get attention on this?
You want me to upgrade to another payment tier? I need this working in prod!
jake
Apologies but this looks like an issue with the application level code. Due to volume, we can only answer platform level issues hereI've made this thread public so that the community might be able to help
2 months ago
After much frustrating testing and generating a new project, suddenly everything was working.
That is, until I changed the private network url
. Now it says:
Error: getaddrinfo ENOTFOUND maxistory.railway.internal
code: 'ENOTFOUND',
syscall: 'getaddrinfo',
hostname: 'maxistory.railway.internal'
}
I tried to revert it to the previous name redis-qjp0
and now that is dead too.
When connecting in terminal I get:
I/O error
I also see in the Railway UI: Unable to connect to the database via the public network:
On my first attempt, I changed this before testing it the first time, but now I see that the change itself was causing the problem.
Now that I've verified it's a "PLATFORM LEVEL ISSUE", can you finally please help me?
2 months ago
I think I've figured it out.
When using the "Create" button to make a single project item (which I suppose now is just a container) I was under the impression that the Redis DB was able to run alongside my project when I did railway up
. By doing that, even though it still showed as REDIS and all the connection info in the Railway UI, I guess I was overwriting the entire container, which broke the connection.
I'm an experienced dev and this was still confusing for me, so either do one of the following:
1. Warn the user they are overwriting an instance when pushing something that is not of the same type via railway up
2. Make this cleaner in the UI AND also remove the existing instance type (Redis logo, db connection info, etc) if they do actually overwrite it with a new project. IE: If they replace a Redis instance with an Express server, the container type should change.
Also, please have better assistance to your community. This was painful, and in the time Jake replied with a nothing burger of a response, he could probably have seen the issue by looking at my project.