What is dynamic cluster management?
Dynamic cluster management is a method of automatically adjusting operational characteristics of a distributed system (cluster), commonly a distributed database like Aerospike. Dynamic cluster management ensures a uniform distribution of data, the management and adjustment of associated metadata and indexes, and automatically adjusting transaction workload. This makes capacity planning and cluster scaling (up or down) decisions precise and simple for Aerospike clusters.
Dynamic cluster management in a distributed system enables you to define cluster members and to create rules to automatically define them. For example, you can create a membership policy that automatically chooses nodes to host dynamic cluster members. It also enables you to easily stop or start server instances as necessary. If it’s operating automatically, then the server instances start and stop automatically. Or, if the dynamic cluster is in supervised mode, then the administrator will be advised to start and stop servers at a specific time. When it’s in manual mode, then the administrator makes the decision about when to stop and start.