Do custom event queues give better performance

[peg612]

I am working on building out a process that is going to be event driven. Has anybody created custom event queues to help ensure the main queue does not get backed up?

I found this article below saying the default queue had a processor running on every active node to help but I am wondering if I am able to create a custom queue that will do the same or do I even need to worry about it.

Improved events processing in Fuji and later releases

[bernyalvarado]

Hi Paul,

We have used the default ecc_queue extensively and have never face any issues using it for other customizations/applications.

Thanks,

Berny

[Valor1]

Paul,

I'm implementing this right now -- you have to specifically design your custom queues to be discrete.

Example -- If you create an additional queue processor, it will pull all the events assigned to it in the "ready" state. So, if you want your processing job to "multi-thread" across multiple background queues, you'll have to create additional mulitple custom queues and figure out how to assign your jobs across the lot of them.

Most of the time, I'd say not to worry about it. If you'd like to know why I am worrying about it, shoot me an email. 🙂

[sergiu_panaite]

Hi Paul,

Custom queues are useful usually when you have events that take time to be processed or you have sudden floods of events that can potentially impact default queue processing. I've seen customers for example with custom queues for events that needed to do an outbound call when they were processed, and that made their processing time to be higher than regular events. In cases like this it is probably a good idea to have a separate events queue. And yes, you can configure the events process job for a custom queue in the same way as the default one (job needs to be marked as "All active nodes" and it will spawn a job per node).

Hope this help, if you have additional questions related to this , let me know.