20-08-2012, 12:15 PM
Decomposing Workload Bursts for Efficient Storage Resource Management
Decomposing.doc (Size: 28.5 KB / Downloads: 26)
ABSTRACT
The growing popularity of hosted storage services and shared storage infrastructure in data centers is driving the recent interest in resource management and QoS in storage systems. The increasing complexity of storage management and the economic benefits of consolidation are driving storage systems toward a service-oriented paradigm, in which personal and corporate clients lease space and access bandwidth on shared storage servers The bursty nature of storage workloads raises significant performance and provisioning challenges, leading to increased resource requirements, management costs, and energy consumption. We present a novel workload shaping framework to handle bursty workloads, where the arrival stream is dynamically decomposed to isolate its bursts, and then rescheduled to exploit available slack. We show how decomposition reduces the server capacity requirements and power consumption significantly, while affecting QoS guarantees minimally. We present an optimal decomposition algorithm RTT and a recombination algorithm Miser, and show the benefits of the approach by evaluating the performance of several storage workloads using both simulation and Linux implementation.
EXISTING SYSTEM
The increasing complexity of storage management and the economic benefits of consolidation are driving storage systems toward a service-oriented paradigm, in which personal and corporate clients lease space and access bandwidth on shared storage servers. In a typical setup, Service Level Agreements (SLAs) between the service provider and clients stipulate guarantees on throughput or latency, for rate-controlled clients. The service provider must provision sufficient resources to meet these performance guarantees based on estimates of the resource demands of the individual clients, and the aggregate capacity requirements of the client mix. The runtime system must isolate the clients to avoid interference, and schedule their requests appropriately.
PROPOSED SYSTEM
In this paper, we present a novel approach to improve client performance and reduce resource provisioning at the server. This results in more predictable behavior, and significantly lower resource requirements