The pizza counter analogy makes this so much clearer than most distibuted system docs. What I find intresting is how Hazelcast handles partition rebalancing when you add new nodes, becuase most frameworks struggle with dat movement without impacting live traffic. Have you tested how it performs when you scale from 3 to 10 nodes under heavy load?
The pizza counter analogy makes this so much clearer than most distibuted system docs. What I find intresting is how Hazelcast handles partition rebalancing when you add new nodes, becuase most frameworks struggle with dat movement without impacting live traffic. Have you tested how it performs when you scale from 3 to 10 nodes under heavy load?
Yes, it works very well, with out pausing current process. It really good in scale up and scale down..