Yes, we were streaming for 24h about Liferay DXP and Liferay Portal 7.3 (and previous versions). All kinds of technical information that might be of interest for you. You might have missed it, but luckily we have it all recorded - and the recordings are available for you:
First I'll offer a brief introduction about what Liferay Cluster means, its basic concepts. I'll explain how do nodes get synchronized at background and which libraries are involved in communication (JGroups) and scheduling (Quartz).
After that I'll explain how to diagnose the cluster current status, how to check current scheduled jobs and understand the different messages that can be obtained due to unexpected situations.
Last, I can show some ways to recover from exceptional states where jobs are lost without the need of stopping nodes.
Basically it would be based on the post I've written: https://liferay.dev/blogs/-/blogs/how-to-diagnose-and-recover-liferay-cluster, but if I've time I can also show how to schedule a PERSISTED task since there are not enough examples about it.
Specially applicable to those who have to deal with cluster stuff, either administrators or developers, that want to get a deeper knowledge about how does Liferay handles cache synchronization or job scheduling, and how they can manage some of the most common failures or understand the reason of why certain states can be reached.