Quality-Elasticity: Improved Resource Utilization, Throughput, and Response Times Via Adjusting Output Quality to Current Operating Conditions

Abstract

This work addresses two related problems for online services, namely poor resource utilization during regular operating conditions, and low throughput, long response times, or poor performance under periods of high system load. To address these problems, we introduce our notion of qualityelasticity as a manner of dynamically adapting response qualities from software services along a fine-grained spectrum. When resources are abundant, response quality can be increased, and when resources are scarce, responses are delivered at a lower quality to prioritize throughput and response times. We present an example of how a complex online shopping site can be made quality-elastic. Experiments show that, compared to state of the art, improvements in throughput (57% more served queries), lowered response times (8 time reduction for 95 th percentile responses), and an estimated 40% profitability increase can be made using our quality-elastic approach. When resources are abundant, our approach may achieve upwards of twice as high resource utilization as prior work in this field.

The 6G Computing Continuum (6GCC): Meeting the 6G computing challenges

“…The system would synthesize the intent into a quality elastic implementation that can be deployed to and function where it deemed needed by the system orchestrator. The applications need to be both extremely resilient and quality-elastic [17], i.e., dynamically adapt to prevailing resource availability wherever they happen to reside in the infrastructure without failures. For applications with timecritical feedback control, missing computation deadlines can be catastrophic to the system under control [18].…”

“…In particular, we plan to identify application-level and applicationdependent requirements, which for a generic application are specific to the domain, but can be translated into a requirement for the infrastructure and its real-time characteristics. Also, we propose to address applications' quality-elasticity [17]. Rather than failing when resources are scarce, a quality-elastic application adapts the quality of its output to the best of its ability given the resources it is allocated.…”

Tigris: a DSL and Framework for Monitoring Software Systems at Runtime

“…The remaining 33 surveyed approaches (51.5%), e.g. [61,62,63,64,65,66], do not even mention the impact of monitoring and gathering data from the managed system. This could raise questions of their practical feasi-378 bility because even in cases where a small amount of data is collected during system monitoring, like execution time and identification of the event [60,67], there is an impact on the memory consumption and processing time of the system.…”