Load Balancing Strategy

Generally, the load balancing method can be classified into two types: static load Balancing algorithm and dynamic load balancing algorithm.   Static load Balancing algorithms include: polling, ratio, priority. ♦         Polling (Round Robin): Sequential loops will request that each server be connected in a sequential loop. When one of the servers fails, do not participate in the next polling until it is back to normal. ♦         ratio (Ratio): Assigns a weighted value to each server, which is proportional to the ratio and assigns the user's request to each server. When one of the servers fails, it does not participate in the next user request assignment until it is back to normal. ♦         Priority (Priority): Group all servers, define priority for each group, and request assigned to the highest priority server group (within the same group, A polling or ratio algorithm is used to assign a user's request, and requests are sent to the server group of the secondary priority when all servers in the highest priority are faulted. In this way, it actually provides a way for the user to warm up.   Dynamic load balancing algorithms include: Minimum connection number, fastest response speed, observation method, predictive method, dynamic performance allocation, dynamic server replenishment, service quality, service type, rule pattern.  ♦         Minimal connection (least Connection): delivers new connections to servers that have minimal connection processing. When one of the servers fails, it takes it out of the server queue and does not participate in the next user request assignment until it is returned to normal. ♦         fastest Mode (fastest): delivers connections to those servers that respond fastest. When one of the servers fails, it takes it out of the server queue and does not participate in the next user request assignment until it is returned to normal.  ♦         observation Mode (observed): Number of connections and response time select the server for the new request based on the best balance of these two items. When one of the servers fails, take it out of the server queue and do not participate in the next userThe requested assignment until it returns to normal. ♦          Predictive Mode (predictive): Use the collected server's current performance metrics for predictive analysis, Select a server in the next time slice, its performance will reach the best server corresponding user's request. ♦          Dynamic RATIO-APM: The Collected application and application server performance parameters, dynamic adjustment of traffic distribution.  ♦          Dynamic Server supplement (Active Server Act.): When the number of failures in the primary server farm is reduced, Dynamically replenish the backup server to the primary server farm.  ♦          Quality of Service (QoS): Allocate data streams at different priority levels.  ♦          service Type (ToS): Allocates data streams by different service types (identified in the type of field).  ♦          rule mode: For different data flow settings guidance rules, users can edit the flow distribution rules, These rules are used to implement the guidance control of the data flow through.