130 likes | 295 Views
LOAD BALANCING IN A CENTRALIZED DISTRIBUTED SYSTEM. BY ANILA JAGANNATHAM ELENA HARRIS. OUTLINE. PROBLEM STATEMENT ARCHITECTURE LOAD BALANCING METHODS EXPERIMENTAL RESULTS CONCLUSION. PROBLEM STATEMENT.
E N D
LOAD BALANCING IN A CENTRALIZED DISTRIBUTED SYSTEM BY ANILA JAGANNATHAM ELENA HARRIS
OUTLINE • PROBLEM STATEMENT • ARCHITECTURE • LOAD BALANCING METHODS • EXPERIMENTAL RESULTS • CONCLUSION
PROBLEM STATEMENT To study different load balancing methods in order to improve CPU utilization and overall execution time of requests in a centralized distributed system with an environment that requires extensive calculations.
ARCHITECTURE • MASTER – SLAVE ARCHITECHTURE • CENTRALIZED CLIENT ACTS AS THE MASTER MACHINE • SERVERS ACT AS SLAVES • SOCKETS ARE USED FOR COMMUNICATION SERVER SERVER SERVER CLIENT
LOAD BALANCING APPROACHES • RANDOM LOAD BALANCING Requests are routed to servers at random. Used for homogeneous systems. • ROUND ROBIN LOAD BALANCING Simple and no overhead Load may not be evenly balanced in the system • CPU UTILIZATION BASED LOAD BALANCING Client queries CPU utilization of each of the servers and chooses the least loaded server to assign incoming user request.
CPU UTILIZATION METRIC • Client requests statistics from /proc/stat file maintained by each server to estimate current CPU utilization. • Statistics from /proc/stat: – Total time spent by the CPU on all processes – Total number of processes running. • Average time per process, T / P is used as metric T – Difference in CPU time at two time points P – Difference in number of processes at two time points • The larger T / P, the less loaded a CPU is.
USER APPLICATIONS • Four applications are chosen to simulate user requests. • Run time of these applications is 8 min , 3 min, 30 sec and 3 sec respectively. • Client has dual functionality - Generates user requests - Performs load balancing.
EXPERIMENTAL RESULTS T/P TIME PERIODS RANDOM LOAD BALANCING
T/P TIME PERIODS
T/P TIME POINTS
EXPERIMENT -II In this experiment, 1000 tasks were distributed. From the graph, average time/process using the proposed method changes less drastically.
CONCLUSION LOAD BALANCING USING CURRENT CPU AVERAGE TIME/PROCESS RESULTS IN BETTER OVERALL CPU UTILIZATION AS COMPARED TO RANDOM OR ROUND ROBIN METHODS.