Hi all, I’m running n8n in standard mode with a concurrency limit set to 50 (N8N_CONCURRENCY_PRODUCTION_LIMIT=50). However, the workflow seems to only run up to 10 executions concurrently, even though the throughput is much higher.
The message is purely informational; it’s simply confirming that the limit for concurrent executions is set to 50, which is correct since I configured it using the environment variable N8N_CONCURRENCY_PRODUCTION_LIMIT.
The issue is that, even when I trigger the MQTT node with 50+ requests simultaneously, I’m unable to fully utilize the available concurrency. The current active executions never exceed 10, whereas I would expect them to reach 50.
Same setup here with N8N_CONCURRENCY_PRODUCTION_LIMIT=80 and it only shows 10 in the execution screen. I am certain I have an external system hitting carder that 10 and these take around 30 seconds to finish. Not sure if 1) the execution screen is the one that lags behind (because i seldom see queued executions in the execution screen) or 2. it is really only letting 10 concurent executions running simultaneously.
Can anybody give us ideas on how to validate this?
Ok, it looks like this is a frontend issue where only up to 10 executions are displayed, but it doesn’t seem to affect actual concurrency.
In the execution_entity table, I can see the correct number of executions with the status ‘running’, which correctly aligns with the N8N_CONCURRENCY_PRODUCTION_LIMIT setting.
