Providing SLO Compliance on NVMe SSDs Through Parallelism Reservation

Abstract

Non-Volatile Memory Express (NVMe) is a specification for next-generation solid-state disks (SSDs). Benefited from the massive internal parallelism and the high-speed PCIe bus, NVMe SSDs achieve extremely high data transfer rates, and they are an ideal solution of shared storage in virtualization environments. Providing virtual machines with Service Level Objective (SLO) compliance on NVMe SSDs is a challenging task, because garbage collection activities inside of NVMe SSDs globally affect the I/O performance of all virtual machines. In this study, we introduce a novel approach, called parallelism reservation, which is inspired by the rich internal parallelism of NVMe SSDs. The degree of parallelism stands for how many flash chips are concurrently active. Our basic idea is to reserve sufficient degrees of parallelism for read, write, and garbage collection operations, making sure that an NVMe SSD delivers stable read and write throughput and reclaims free space at a constant rate. The stable read and write throughput are proportionally distributed among virtual machines for SLO compliance. Our experimental results show that our parallelism reservation approach delivered satisfiable throughput and highly predictable response to virtual machines.

Parallelism-Aware Channel Partition for Read/Write Interference Mitigation in Solid-State Drives

“…Interference within the workload has not been addressed Huang et al [13] Lv et al [14] Allocate different types of requests into different partitions…”

“…Performance isolation schemes exploit the fact that the main cause of R/W interference is the sharing of flash memory resources among the different types of requests. The objective of performance isolation techniques [12,13,16,17] is to provide predictable latency by preventing interference between tenants sharing an SSD. For example, Huang et al and Kim et al deployed a similar approach to data allocation to avoid performance degradation caused by the garbage collection (GC) of other tenants [12,13].…”

“…The objective of performance isolation techniques [12,13,16,17] is to provide predictable latency by preventing interference between tenants sharing an SSD. For example, Huang et al and Kim et al deployed a similar approach to data allocation to avoid performance degradation caused by the garbage collection (GC) of other tenants [12,13]. Although performance isolation schemes reduce inter-tenant interference, they overlook R/W interference within a single workload.…”

Partitioned Real-Time NAND Flash Storage

“…Huang et al [9] exploited the internal parallelism in SSDs by reserving banks for servicing read and write requests and performing garbage collection to guarantee stable read and write throughput. Partial garbage collection is used, where each step is either a page copy or a block erasure, so each page read is potentially blocked by a partial step.…”