Hey there! We've found the following might help you get unblocked faster:

• 🧵 Redis Connection ERROR

• 🧵 N8N crash after deploy

• 🧵 N8N from template keeps failing due to Redis connection error

• 🧵 redis connection error - unable to connect

If you find the answer from one of these, please let us know by solving the thread!

Sorry it is unsolved

Also happening here!! It has been working with no error for months

Is this a connection to N8N for you too?

We have the same issue with n8n and redis here on Railway. And it also started about 2-3 days ago. Before, it worked well for months.

We've got like that in logs

[Redis client] read ECONNRESET

Queue errored

read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/20s)

Recovered Redis connection

[Redis client] read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/20s)

[Redis client] read ECONNRESET

Queue errored

read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/20s)

Recovered Redis connection

[Redis client] read ECONNRESET

Queue errored

read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/20s)

Recovered Redis connection

[Redis client] read ECONNRESET

Queue errored

read ECONNRESET

Unable to connect to Redis after trying to connect for 20s

And workers crashes after that.

We have the same issue with our n8n setup. Mostly the worker node crashes, however sometimes both the worker and primary nodes crashes.
The same error messages as above occur.

We also have the same issue for both workers and primary instances. How to resolve this?

For us, the issue starts every day at midnight

Exactly! Always at midnight!

Related topics with the same issue
- https://station.railway.com/support/redis-disconnect-0b8611d8
- https://station.railway.com/questions/worker-is-failing-263de60c

I'm having the same issue! I have 2 projects, and the same worker is failing

i changed the variables in worker to this values, it's crashing, but less frequent:

////////////////////////////////////////////////
DB_POSTGRESDB_DATABASE="${{Postgres.POSTGRES_DB}}"

DB_POSTGRESDB_HOST="${{Postgres.PGHOST_PRIVATE}}"

DB_POSTGRESDB_PASSWORD="${{Postgres.POSTGRES_PASSWORD}}"

DB_POSTGRESDB_PORT="${{Postgres.PGPORT_PRIVATE}}"

DB_POSTGRESDB_USER="${{Postgres.POSTGRES_USER}}"

DB_TYPE="postgresdb"

ENABLE_ALPINE_PRIVATE_NETWORKING="true"

EXECUTIONS_MODE="queue"

N8N_BLOCK_ENV_ACCESS_IN_NODE="false"

N8N_ENCRYPTION_KEY="${{Primary.N8N_ENCRYPTION_KEY}}"

N8N_ENFORCE_SETTINGS_FILE_PERMISSIONS="true"

N8N_EXECUTIONS_MODE="queue"

N8N_REDIS_HOST="${{Redis.REDISHOST}}"

N8N_REDIS_PASSWORD="${{Redis.REDIS_PASSWORD}}"

N8N_REDIS_PORT="${{Redis.REDISPORT}}"

N8N_REINSTALL_MISSING_PACKAGES="true"

N8N_ROLE="worker"

N8N_RUNNERS_ENABLED="true"

NODE_OPTIONS="--max_old_space_size=8192"

OFFLOAD_MANUAL_EXECUTIONS_TO_WORKERS="true"

QUEUE_BULL_REDIS_DUALSTACK="true"

QUEUE_BULL_REDIS_HOST="${{Redis.REDISHOST}}"

QUEUE_BULL_REDIS_PASSWORD="${{Redis.REDIS_PASSWORD}}"

QUEUE_BULL_REDIS_PORT="${{Redis.REDISPORT}}"

QUEUE_BULL_REDIS_USERNAME="${{Redis.REDISUSER}}"

////////////////////////////////////////////////

and in primary is this values:

////////////////////////////////////////////////

DB_POSTGRESDB_DATABASE="${{Postgres.POSTGRES_DB}}"

DB_POSTGRESDB_HOST="${{Postgres.PGHOST_PRIVATE}}"

DB_POSTGRESDB_PASSWORD="${{Postgres.POSTGRES_PASSWORD}}"

DB_POSTGRESDB_PORT="${{Postgres.PGPORT_PRIVATE}}"

DB_POSTGRESDB_USER="${{Postgres.POSTGRES_USER}}"

DB_TYPE="postgresdb"

ENABLE_ALPINE_PRIVATE_NETWORKING="true"

N8N_BLOCK_ENV_ACCESS_IN_NODE="false"

N8N_EDITOR_BASE_URL="https://${{RAILWAY_PUBLIC_DOMAIN}}"

N8N_ENCRYPTION_KEY="***********************************************"

N8N_ENDPOINT_REST="workflow/rest"

N8N_ENFORCE_SETTINGS_FILE_PERMISSIONS="true"

N8N_EXECUTIONS_MODE="queue"

N8N_HOST="**************************"

N8N_LISTEN_ADDRESS="::"

N8N_QUEUE_BULL_REDIS_HOST="${{N8N_REDIS_HOST}}"

N8N_QUEUE_BULL_REDIS_PASSWORD="${{N8N_REDIS_PASSWORD}}"

N8N_QUEUE_BULL_REDIS_PORT="${{N8N_REDIS_PORT}}"

N8N_REDIS_HOST="${{Redis.REDISHOST}}"

N8N_REDIS_MAX_RETRIES="3"

N8N_REDIS_PASSWORD="${{Redis.REDISPASSWORD}}"

N8N_REDIS_PORT="${{Redis.REDISPORT}}"

N8N_REDIS_RETRY_DELAY="2000"

N8N_REDIS_TIMEOUT="60000"

N8N_REINSTALL_MISSING_PACKAGES="true"

N8N_RUNNERS_ENABLED="true"

NODE_OPTIONS="--max_old_space_size=8192"

OFFLOAD_MANUAL_EXECUTIONS_TO_WORKERS="true"

PORT="5678"

REDIS_COMMAND_TIMEOUT="60000"

TZ="***************************"

WEBHOOK_URL="**********************"

But the Railway should fix this

Who guarantees that this isn't a problem with the N8N version? Maybe the problem is with them.

Because my n8n was up for like 4 months without any updates, and it was working fine, nothing changed (probably only railway), and I even tried to update just to be sure, as you said, but the issue was still happening. There is an error in the connection between services that Railway coordinates

Hey guys, if all of us have the same issue:
- problem with n8n instance redis lost connection
- it started a few days ago, only on Railway
- we expereince it every day near midnight
- it doesn't depend on n8n version

It looks like we need Railway support help. Is there a way to tag them here?

Exactly I created a bug ticket in discord. Maybe you guys can do the same? https://discord.com/channels/713503345364697088/1414958654234296381/1414958654234296381

I'm experiencing the same issue; I thought it was a plan problem, but I upgraded to Pro, and it's still the same.

It just failed again, and it is not midnight

Same for me

Yes, the same thing. It failed again, 16:11 UTC.

Yes, it happened to mine as well, it just crashed a few minutes ago.

Mine too. It seems to be a general glitch.

Same thing has been happening to me as well i am hosting n8n on render ans postgres on railway and error 503 db connection timeout keeps happening for one reason or another. i tried everything, but it did not resloved i am thinking of hosting postgres on render as well what do you guys think

Crash a few minutes ago(
How do you think is it a good idea to change hosting only for redis instance?

How complicated is it to transfer the data to another server from N8N? With all the credentials.

I had the same issue. This started after that update the Railway did in last week.

yes thats probably the cause

Here the same problem.

Same problem here - has anyone figured out how to move from the bitnami/redis image to a different one without loosing the data in their redis DB?

same here. Any idea on fix steps?

Queue errored

read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection

[Redis client] read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

[Redis client] read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection

[Redis client] read ECONNRESET

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection

[Redis client] read ECONNRESET

Queue errored

read ECONNRESET

Unable to connect to Redis after trying to connect for 10s

Exiting process due to Redis connection error

Same issue here

I'm also having the same issue. Anyone found a solution that works? Happening more now to me.

I tried changing the Redis image from Bitnami to the new one forked by Railway, but it didn’t work either

Well that's discouraging. I upgraded thinking my account was the issue. Hope we can find a solution soon.

I have changed the Restart Policy in the settings for N8N Primary and Worker to "Always", so after it crashes again, it will restart instead of keeping down. It might help while we wait for a solution.

You can find the setting here.

Sweet, thanks. Just did that.

Same here.. N8N worker primary crashing every 4 hours....never had issues with it .. were doing a lot of volume , even tried 2 instances for workers full 32cpu and 32gb ram and still crashing... did you get it to work>

literally same here... such a headache.

has anyone found at least a workaround?

For now turn the restart policy for worker and primary to "Always" that would help if it crashes again

Same issue here....

anyone tried using the new source image for redis on n8n? railwayapp/redis:8.2

If I update will it erase my WF or credentials?

I didn't experience the issue today after 02:00 UTC. What about you guys?

same but I added a few variables taken from the newest n8n template w worker. but redis source still the old one even on the new templates for n8n. bitnami instead of railwayapp/redis:8.2

Anyone tried using that source for redis?

If you are experiencing disconnects with Redis, can you please ensure that you are connecting to it over the private network

You should be able to do something like this

• QUEUE_BULL_REDIS_HOST="${{Redis.REDISHOST_PRIVATE}}"

• QUEUE_BULL_REDIS_PORT="${{Redis.REDISPORT_PRIVATE}}"

• N8N_REDIS_HOST="${{Redis.REDISHOST_PRIVATE}}"

• N8N_REDIS_PORT="${{Redis.REDISPORT_PRIVATE}}"

For both the Worker and Primary

Make sure to backup the variables before making this change though

it is not working

i changed them like this

They crashed instantly

I have these by default

• QUEUE_BULL_REDIS_HOST="${{Redis.REDISHOST_PRIVATE}}"

• QUEUE_BULL_REDIS_PORT="${{Redis.REDISPORT_PRIVATE}}"

but not these

• N8N_REDIS_HOST="${{Redis.REDISHOST}}"

• N8N_REDIS_PORT="${{Redis.REDISPORT}}"

these go on both worker and primary??

doesnt have to do with the source image from bitnami? i never had issues with n8n crashing at all and now crashes all the time for the past 5 days or so ?

how can we fix this?

The underlying Redis images do not have any behaviour that is different.

Unfortunately I do not know anything about n8n specifics. It might be that it does not support the private network? I found these related threads https://station.railway.com/questions/n8n-crash-enotfound-redis-785768fa, https://station.railway.com/questions/n8n-deployment-crashed-error-0f992914.

Will have to rely on the community for support here

How can you suggest using a Redis private host and port when another employee is also mentioning that it's not supported in the other thread?

Also, I think this is an issue with the Railway. I didn't do any updates on n8n; I didn't change anything. The deployment was there for about four months, and it was working. Suddenly, it started breaking, and it's not just me; it's a bunch of people who are facing these issues.

Sorry, I don't wanna sound mean or something, but I'm not on a pro account, but I'm paying. Maybe it's not a lot for you guys, but I'm still paying, and a lot of us are facing this issue. People who are actually pro are having the same issue. As you can see in the thread, some people even updated their accounts just to try to fix this because they thought that was a problem, and now we can't have a solution, and this is impacting our services and everything that we are providing for our customers.

Your n8n workflows are failing because Redis is asking for authentication,

but n8n doesn't have the password configured.

What's going on: Railway automatically sets up Redis with authentication

enabled, but the connection credentials aren't being passed to your n8n

service.

Quick fix:

Head to your Railway dashboard and find your Redis service - look in the

Variables tab for the connection details. Then add these environment

variables to your n8n service:

DB_REDIS_HOST=${{Redis.REDISHOST}}

DB_REDIS_PORT=${{Redis.REDISPORT}}

DB_REDIS_PASSWORD=${{Redis.REDISPASSWORD}}

Railway will automatically inject the Redis password using that

${{Redis.REDISPASSWORD}} reference.

After you add those variables and restart the service, n8n should connect

to Redis without issues and your workflows will stop crashing.

This is pretty common when services get deployed separately - they just

need the proper connection configuration to talk to each othe

The reply in the screenshot is from 8 months ago. This template, https://railway.com/deploy/n8n-with-workers, looks to now use the private network for Redis so I think something has changed.

I am just making suggestions purely on the Railway configuration. I've personally never used n8n or know anything about its internals or how it works. The few errors shared here all look like application errors outside of Railways control.

I didn't do any updates on n8n

Redeploying on Railway will pull in the latest n8n version which might include an update.

Apologizes that we can't be more helpful with n8n issues. There are many users currently running n8n with success and can hopefully provide help here (and receive a bounty!). The linked n8n-with-workers template also successfully deploys, so you might be able to look at that for a configuration comparison (I recommend backing up your n8n before!).

I already updated and tried to update n8n, but it didn't work. I'm using the same template you're referring to, so I'm surprised it's not working. A lot of us are experiencing this issue; I can understand if it was just me, but I believe many users are having the same problem. Already updated, I already tried to update n8n, and it didn't work. And actually, I'm running using the same template that you are pointing to. But that's why I'm surprised that it is not working. And a lot of us, I mean, I could understand if it was just me, but I think there are a lot of users who were having that issue

Thank you for your answer, but I have a question. If that's the issue, why is it connecting using the public network, and how it was working for so long, and suddenly it works, but every 4 hours or something, it crashes? Will these changes help me?

We fixed a bug with public networking issues recently that was causing random connection drops.

Are you still experiencing this issue after switching to the private network?

No we haven’t experienced the issue since yesterday

Hello everyone! If anyone is still experiencing n8n issues, please read this.

Redis Connection Problems

1. In your n8n primary and n8n worker(s), add the following variables:

QUEUE_BULL_REDIS_DUALSTACK=true
N8N_REDIS_HOST=${{Redis.REDISHOST_PRIVATE}}
N8N_REDIS_PORT=${{Redis.REDISPORT_PRIVATE}}

2. Redeploy n8n primary and n8n worker(s)

Note that the variables may be slightly different depending on your setup/what you have deployed. You can also check out our docs on Reference Variables for the exact syntax if you're unsure.

This will ensue your connection goes over the private network, instead of the public network which may have been unstable over the past few days as we were fixing a bug with it.

Please also note that the emails we sent regarding bitnami redis update is unrelated to this.

Postgres Connection Problems

The logs you're seeing about "collation version mismatch" is a warning and has no impact on your database. However, you can resolve this by connecting to your Postgres instance using a SQL client such as TablePlus and running:

ALTER DATABASE railway REFRESH COLLATION VERSION;
REINDEX DATABASE railway;

The templates you've deployed by default, from what I can see, connects to Postgres over the private network. So there is no configuation changes required here. If you're still having issues with Postgres connections, please create a new thread and link to your Postgres service.

Image Auto-Update

1. In your n8n primary, add the following variable:

N8N_RUN_MIGRATIONS=true

2. Redeploy your n8n primary and n8n worker(s)

This will run database migrations for n8n, which is required when a version change results in an update of n8n's schema. We automatically update images in accordance with industry best practices to ensure systems remain current and secure. n8n recommends auto-updating as well.

As a general note, to prevent issues like this from happening in the future, I recommend using the n8n-with-workers template. It is created and maintained by Jack, a valuable community member we work closely with.

---

And as a closing note...

Apologies for the poor experience here. Please note that we generally don't provide first-party support for third-party software deployed on Railway. To use an analogy, Railway is an "electricity provider" and the software you deploy are "appliances", so it is unfortunately out of our control and purview (i.e. we function as the infrastructure provider rather than the application maintainer).

The connection issues stemmed from us and caused a lot of pain and confusion in here, so we're making an exception to intervene in this case. I hope this helps, and again, please accept my sincerest apologies for all the issues you've faced as a result of this.

Best,

Ray from Railway

worked, tnx

Ray, Much appreciated. Thanks for jumping into this and helping us resolve this massive issue.

Sometimes the electricity provider knows something appliance people often don't, so thanks again.

Hi Rey,
thanks for your support.

It looks like everything is fine now!