Executions in queue not executed

Questions
1

Self-hosted N8N 1.85.4

Got a working workflow triggered by webhook.

Attempted to switch to queue mode for better performance

Setup:
Main N8N in Docker on VPS
Worker N8N in Docker on VPS
Redis on VPS
PostgreSQL on VPS

Ubuntu 22

Both N8N instances within containers can connect to Redis and PostgreSQL (this is clear from logs)

Executions stay in ‘queued’ status.

Tried numerous troubleshooting, nothing helps.

Apparently, PostgreSQL is populated with execution in ‘new’ status, but Redis does not (no n8n:queue, but there are other n8n-related data in Redis - to be honest, not sure how it is supposed to look in reality, the required ‘n8n:queue’ is suggestion by chatgpt, it sometimes invents stuff…).

Redis is definitely operating fine, also used by other application on the server.

Both main instance and worker have ‘Querying database for waiting executions’ in logs.

Any help or direction is really appreciated

Thank you

2

Hi,

the default name of the queue would be something like bull:jobs

You can see an example here

E.g Waiting jobs would/should be bull:jobs:wait

Reg,
J.

3

I checked Redis queue, apparently it does have execution object.

Here is log from the main N8N instance:

a1
a11801×692 185 KB

Log from worker

a2
a21815×230 52 KB

Probably this is the execution object in Redis

a3
a31821×408 23.1 KB

Nothing happens…

4

Hi,

Maybe you can compare your setup with the n8n-hosting templates and verify whether it works with the versioning you are testing.

There is a docker-compose file which has all setting required for a N8n postgres and worker scenario

Reg,
J.

5

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.