Question

49
Views
Close popover
Cloe Walker (CloeW938)
IT Solution Service

IT Solution Service
SG
CloeW938 Member since 2018 151 posts
IT Solution Service
Posted: September 7, 2020
Last activity: September 8, 2020

How does system identify multiple nodes as the same cluster?

Hazelcast is a purely distributed and decentralized cluster. Hence there is no master node concept and each member joins a certain cluster somehow. My question is, how does each member identify each other as the same cluster? I know if you specify members ip address in the prconfig.xml file it works but I want to know how system handles this without prconfig.xml, using an empty table (pr_sys_statusnodes). I find "pxClusterID" in the table but is this the one that groups nodes? If so, how is this determined?

If I ask differently, how can force "Split-Brain Syndrome" to occur (https://community.pega.com/knowledgebase/articles/system-administration/split-brain-syndrome-and-cluster-fracturing-faqs)? I tried starting Tomcat at multiple nodes at the same time but it doesn't seem to happen. Why?

Thanks,

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

Pega Platform 8.4 Data Integration