KeyValueServe†: Design and performance analysis of a multi-tenant data grid as a cloud service

摘要

Distributed key-value stores have become indispensable for large-scale cluster applications.rnMany cloud services have deployed in-memory data grids for their enterprise infrastructures andrnsupport multi-tenancy services. However, most services do not offer fine-grained multi-tenantrnresource sharing. To this front, we present KeyValueServe, a low overhead cloud service with featuresrnaiding resourcemanagement. Results based on Hazelcast, a popular open source data grid,rnindicate thatKeyValueServe can efficiently provide services to tenants without degrading performance.rnProviding consistent performance to all tenants for fluctuating workloads is still difficult.rnPerformance problems occur at scale with diverse tenant requirements. To address this, the paperrnprovides insights to contention and performance bottlenecks. Through experimental analysis,rnwe uncover scenarios of performance degradation and demonstrate optimized performance viarncoalescing multiple clients' requests. Our work indicates that a Hazelcast cluster can get congestedrnwith multiple concurrent connections when processing client requests, resulting in poorrnperformance.KeyValueServe can reduce the number of parallel connections maintained for clientrnrequests, resulting in improved performance.