Constant throughput
To achieve constant throughput is a good success factor. What if you start experiencing a workload that starts affecting the rest of the members in group replication? It might be a case where your master keeps on accepting additional workload and is lagging behind, after which it might return to an acceptable level before burning out all the resources. Additionally, you can implement a queuing methodology that can prevent you from burning down resources and only allows you to pass on workloads to the members that are predefined based on capacity.
While considering a queuing methodology, you mustn't allow queues grow exponentially. This would impact the end user, as there would be a lag in the data being updated. However, ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access