Redis ENOTFOUND over private network (ipv6) on Railway

[antonlvovych]

Bug Description

When deploying n8n in a queue mode on Railway without publicly exposed Redis, the n8n instances crash with the following error:

[Redis client] getaddrinfo ENOTFOUND redis.railway.internal

This issue occurs when n8n attempts to connect to Redis using Railway’s private network (redis.railway.internal).
The root cause appears to be ioredis not supporting IPv6, which is used for private networking in Railway.

Additional Context

Potential Fix:
According to a Railway employee response, n8n does not support connecting to Redis via the private network due to upstream ioredis misconfigurations (it doesn't resolve IPv6, which is used by Railway in a private network).

If that's the case, it could be fixed by adding an option for ioredis dual stack lookup, which will try to resolve the redis.railway.internal hostname using both IPv4 and IPv6 addresses (A and AAAA records).

• Link to Railway forum: n8n crash ENOTFOUND Redis
• Link to Railway docs: ENOTFOUND redis.railway.internal

Current workaround:
Update Redis connection settings in n8n to use the public Redis URL instead of private (redis.railway.internal).

⚠️ Security Risk: Using the public network exposes Redis unnecessarily, which is not a good practice and should be avoided unless necessary.

To Reproduce

1. Deploy this template to Railway:

2. Check deploy logs of Primary or Worker n8n instance.

Both instances fail with the following error:

[Redis client] getaddrinfo ENOTFOUND redis.railway.internal

Expected behavior

n8n should successfully connect to Redis without encountering ENOTFOUND redis.railway.internal, resolving the hostname via both IPv4 and IPv6.

Operating System

N/A

n8n Version

latest

Node.js Version

N/A

Database

PostgreSQL

Execution mode

queue

[Joffcom]

Hey @antonlvovych,

We have created an internal ticket to look into this which we will be tracking as "CAT-627"

[Joffcom]

Hey @antonlvovych

Thanks for the docs I will pass this on internally for the team to look into, it makes sense to do this however it is more of an enhancement request than a bug so while I should close this and directly it to the support forum it feels like something that should remain here for now.

I just wanted to note that the person that told you this does not work for n8n but railway, I am sure it is just a typo but this should save time for anyone else who looks at this and wonders who it might have been 🙂

[antonlvovych]

@Joffcom yeah, that was a typo — it’s a Railway employee. I just faced this issue today and found this discussion. It looks valid, so it would be great if you could add support for IPv6 networks. PostgreSQL works just fine using an IPv6 private network on Railway.

I see there is already a PR with a fix by @netroy ❤️‍🔥🚀 Can we keep this issue open for now?

[Joffcom]

Hey @antonlvovych ,

Of course, I raised this with Netroy after posting. I just need to review it then we can merge and release it as part of the next release.

[antonlvovych]

@Joffcom @netroy thanks a ton for the super quick upgrade!

[netroy]

Starting next release, you can set the env variable QUEUE_BULL_REDIS_DUALSTACK to true to enable IPv6 support.