Close popover
Cloe Walker (CloeW938)
IT Solution Service

IT Solution Service
CloeW938 Member since 2018 151 posts
IT Solution Service
Posted: June 26, 2020
Last activity: July 6, 2020

What is the general guideline for node type design


We will be using Pega Platform on premise, with no strategic applications like CDH, Marketing, Decisioning, etc. Only Pega Platform. No cloud.

What is the general guideline for node type design.

I understand we can assign each node "BackgroundProcessing", "Search", "Stream", "WebUser", or "RunOnAllNodes". If you assign "RunOnAllNodes", the node will do all the work, but if you assign only one type, the node will handle the specific purpose only.

I understand you may say "yeah, it depends on the volume of Queue Processor, WebUser, etc etc" but what is the general guideline for this design. Pega Community talks a lot about description about Node Type but I can't find a source about realistic design use case.

For example, If we have reasonable number of Agent (Queue Processor), like 10. We are going to have maybe a couple of app servers. In that case, still is it recommended to dedicate each node? Or should we make it all "RunOnAllNodes" because BackgroundProcessing and Search would not be heavy in this case?



***Edited by Moderator: Pallavi to update platform capability tags***  

Pega Platform 8.5 Data Integration