Abstract: The weighted load balancing method on data access nodes ensures the ability to horizontally scale the data access system, the load handling capacity of the system is increased linearly according to the number of data access nodes. The proposed method includes the following steps: step 1: update the routing table, when adding, removing nodes or changing nodes' weight, move virtual nodes from node having decreased number of virtual nodes to node having increased number of virtual nodes; step 2: store old routing table on array Ai and new routing table on array A2; step 3: block access to records that need to be moved; step 4: copy records from old node to node; step 5: perform read/write access using data partitioning method with new routing table A2; step 6: clean duplicated records.

Abstract: The weighted load balancing method on data access nodes ensures the ability to horizontally scale the data access system, the load handling capacity of the system is increased linearly according to the number of data access nodes. The proposed method includes the following steps: step 1: update the routing table, when adding, removing nodes or changing nodes' weight, move virtual nodes from node having decreased number of virtual nodes to node having increased number of virtual nodes; step 2: store old routing table on array A1 and new routing table on array A2; step 3: block access to records that need to be moved; step 4: copy records from old node to node; step 5: perform read/write access using data partitioning method with new routing table A2; step 6: clean duplicated records.