A Multi-Tenant Key-Value SSD with Secondary Index for Search Query Processing and Analysis

Key-Value SSDs (KVSSDs) introduced so far are limited in their use as an alternative to the key-value store running on host due following technical limitations. First, they were designed only for single tenant, limiting of multiple tenants. Second, mainly focused designing indexes primary key-based searches, without supporting various queries using a combination key and non-primary attributes-based searches. This paper proposes Cerberus , Log Structured Merged (LSM) tree-based KVSSD armed with (i) namespace performance isolation tenants multi-tenant environment, (ii) capability processing search queries. Specifically, identifies tenant’s splits large LSM-tree into namespace-specific also manages secondary enable data access fast query processing. With SSD-internal CPU/DRAM resources, supports handles complex that combined computing operations. We prototyped Cosmos+ OpenSSD platform. When there tenants, exhibits up 2.9 × higher read throughput negligible write overhead compared existing KVSSD. shows lower latency by 9.31 query.