Energy Efficient Scheduling of Servers with Multi-Sleep Modes for Cloud Data Center

ABSTRACT

In a cloud data center, servers are always over-provisioned in an active state to meet the peak demand of requests, wasting a large amount of energy as a result. One of the options to reduce the power consumption of data centers is to reduce the number of idle servers, or to switch idle servers into low-power sleep states. However, the servers cannot process the requests immediately when transiting to an active state. There are delays and extra power consumption during the transition. In this paper, we consider using state of- the-art servers with multi-sleep modes. The sleep modes with smaller transition delays usually consume more power when sleeping. Given the arrival of incoming requests, our goal is to minimize the energy consumption of a cloud data center by the scheduling of servers with multi-sleep modes. We formulate this problem as an integer linear programming (ILP) problem during the whole period of time with millions of decision variables. To solve this problem, we divide it into sub-problems with smaller periods while ensuring the feasibility and transition continuity for each sub-problem through a Backtrack-and-Update technique. We also consider using DVFS to adjust the frequency of active servers, so that the requests can be processed with the least power. Our simulations are based on traces from real world. Experiments show that our method can significantly reduce the power consumption for a cloud data center.

EXISTING SYSTEM

To handle the possible peak demand of user requests, servers are always over-provisioned, wasting a lot of energy as a result. Therefore, there is an urgent need to enhance energy efficiency for cloud data centers. The existing work has mainly focused on dynamic voltage frequency scaling (DVFS) and dynamic power management (DPM). The former is to adjust the voltage/frequency of CPU power according to the demand of computing capacity, while the latter reduces the total energy by putting servers into sleep states or turning off idle servers. However, a difficult issue is that the servers cannot process the incoming requests immediately when transiting to active state. There are delays and extra power consumption during the transitions, which have been ignored in the existing work. Besides, modern servers are usually designed with several sleep states, and the sleep states with smaller transition delays consume more power when sleeping.

PROBLEM STATEMENT

• The servers cannot process the incoming requests immediately when transiting to active state.
• There are delays and extra power consumption during the transitions
• Modern servers are usually designed with several sleep states, and the sleep states with smaller transition delays consume more power when sleeping

PROPOSED WORK

In this paper, we propose the AVVMC VM consolidation scheme that focuses on balanced resource utilization of servers across different computing resources (CPU, memory, and network I/O) with the goal of minimizing power consumption and resource wastage. Since the VM consolidation problem is strictly NP-hard and computationally infeasible for large data centers, we propose adaptation and integration of the Ant Colony Optimization (ACO) meta heuristic with balanced usage of computing resources based on vector algebra. Our simulation results show that AVVMC outperforms existing methods and achieves improvement in both energy consumption and resource wastage reduction.

CONCLUSION

In this paper, the problem of scheduling of servers with multi-sleep modes for cloud data centers is investigated . The servers can make transitions between one active state and different sleep states, which involves different sleep power and transition delays for the sleep modes. The proposed AVVMC VM consolidation scheme that focuses on balanced resource utilization of servers across different computing resources (CPU, memory, and network I/O) with the goal of minimizing power consumption and resource wastage. Since the VM consolidation problem is strictly NP-hard and computationally infeasible for large data centers, we propose adaptation and integration of the Ant Colony Optimization (ACO) meta heuristic with balanced usage of computing resources based on vector algebra. Our simulation results show that AVVMC outperforms existing methods and achieves improvement in both energy consumption and resource wastage reduction.