I am connecting n8n to a rabbitMQ Server, everything works fine until the rabbitMQ server is restarted.
N8N doesn’t reconnect or complain about the restart. It looks like the workflow is active but nothing happens.
Deactivating and reactivating the workflow triggers a reconnect which sometimes kills my VM depending on how many rabbitMQ messages are waiting since n8n starts all of them at once.
Unfortunately the rabbitMQ Servers get’s restarted quite often since it’s running in Docker for a different project. Yes I could change to restart only the rails app instead of all containers from the project with every deploy but still it would be great if n8n would reconnect.
The other issue would probably disappear but should maybe also be considered as this could result in some nasty DoS if someone is funny and bombards n8n with a gazillion rabbitMQ messages (it only takes about 20-30 to kill my VM)
Yeah, I guess a reconnection logic like this would be needed.
Hi, is there anything being done in this area or should I post it as github issue?
The current version is actually worse, I can’t stop the workflow once the connection is lost.
The only way to get it working again is by restarting the complete n8n docker container.
ok, looks like there already is an issue on github RabbitMQ trigger doesn't reconnect · Issue #1988 · n8n-io/n8n · GitHub