Hi everyone, I
’ve been working on a custom integration for our Zammad instance to help automate some of our repetitive support ticketing, but I’ve run into a frustrating performance issue. We are trying to push real-time data from a local environment into Zammad via the REST API, but I’m seeing a significant delay in ticket creation whenever my background automation is running at a high frequency.
I actually started looking into how other high-performance environments manage their execution threads after reading a guide at regarding their optimization for rapid data flow. I’ve been trying to see if the logic used for blox fruit scripts to keep execution stable on a local machine could be adapted to my Zammad workflow, as I suspect my local script is hitting a rate limit or causing a memory leak that slows down the Ruby on Rails background workers. You can see the kind of rapid-fire script execution I’m talking about over at , and it made me wonder if Zammad’s “Scheduler” is particularly sensitive to these types of external executors.
I’m also seeing a related problem where the “Elasticsearch” index doesn’t seem to update immediately when a ticket is created via the script, making it impossible for our agents to search for those new tickets for several minutes. Has anyone else noticed if running script-heavy environments alongside a self-hosted Zammad instance causes the “Nginx” buffer to overflow? I’ve already tried increasing my client_max_body_size and adjusting the database.yml pool settings, but the intermittent timeouts persist. Is there a way to prioritize the Zammad API thread so these background executors don’t cause a “Connection Refused” error, or should I be looking into a more robust message broker like Redis to sit between my scripts and the API? I’d really appreciate any advice on keeping the ticketing system snappy without sacrificing my automation tools!