Abstract
Massive data transfers in modern data-intensive systems resulting from low data-locality and data-to-code system design hurt their performance and scalability. Near-Data processing (NDP) and a shift to code-to-data designs may represent a viable solution as packaging combinations of storage and compute elements on the same device has become feasible. The shift towards NDP system architectures calls for revision of established principles. ions such as data formats and layouts typically spread multiple layers in traditional DBMS, the way they are processed is encapsulated within these layers of abstraction. The NDP-style processing requires an explicit definition of cross-layer data formats and accessors to ensure in-situ executions optimally utilizing the properties of the underlying NDP storage and compute elements. In this paper, we make the case for such data format definitions and investigate the performance benefits under RocksDB and the COSMOS hardware platform.
Highlights
Besides substantial data ingestion, yielding an exponential increase in data volumes, modern data-intensive systems perform complex analytical tasks
We prototyped the approach with a simple image processing application, on NoFTL-KV and the COSMOS OpenSSD as real hardware target, and gain performance improvements of up to 33%
We motivate the necessity for data format pushdown in Near-Data Processing (NDP) scenarios
Summary
Besides substantial data ingestion, yielding an exponential increase in data volumes, modern data-intensive systems perform complex analytical tasks. Systems trigger massive data transfers that impair performance and scalability, and hurt resource- and energy-efficiency. These are partly caused by the scarce system bandwidth in combination with poor data locality, as well as by traditional system. A shift towards Near-Data Processing (NDP) and code-to-data allows executing operations in-situ, i.e. as close as possible to the physical data location, leveraging the much better on-device I/O performance. This observation is supported by several trends. The two trends lift major limitations of prior approaches such as ActiveDisks or Database Machines
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
