Containerized In-Storage Processing and Computing-Enabled SSD Disaggregation

TL;DR

DockerSSD introduces a containerized in-storage processing framework on SSDs, utilizing virtualization and lightweight firmware to improve performance and scalability for data analytics and large-scale AI inference.

Contribution

It presents DockerSSD, a novel ISP model combining OS virtualization and firmware techniques for efficient, disaggregated SSD-based data processing.

Findings

Up to 2.0x performance improvement for I/O workloads

7.9x faster distributed LLM inference

Supports large-scale, disaggregated storage pools

Abstract

ISP minimizes data transfer for analytics but faces challenges in adaptation and disaggregation. We propose DockerSSD, an ISP model leveraging OS-level virtualization and lightweight firmware to enable containerized data processing directly on SSDs. Key features include Ethernet over NVMe for network-based ISP management and Virtual Firmware for secure, efficient container execution. DockerSSD supports disaggregated storage pools, reducing host overhead and enhancing large-scale services like LLM inference. It achieves up to 2.0x better performance for I/O-intensive workloads, and 7.9x improvement in distributed LLM inference.