Micron speeds up SSDs with open-source database storage engine

Micron releases an open-source storage engine that optimizes solid-state drives and storage-class memory database throughput, latency and endurance.
Written by Steven Vaughan-Nichols, Senior Contributing Editor

Want to get the most from your database? Then you probably already use Database Storage Engines. Now, Micron Technology, a leading storage company, has announced the first open-source, heterogeneous-memory storage engine (HSE). This is designed for database management systems (DBMS)s to get the most from solid-state drives (SSDs) and storage-class memory (SCM). 

There are many older storage engine designs. Micron claims these hard disk drives -- (HDDs)-based B-Tree and Log-Structured Merge (LSM) Tree engines -- haven't kept up with today's higher performance and reduced latency of SSDs and SCM. HSE, on the other hand, is delivered as an embeddable key-value database. Micron states it's "ideal for developers using all-flash infrastructure." 

Because HSE is now open-source software, DBMS developers can customize or enhance their code for their unique use cases. HSE comes pre-integrated with MongoDB, the most popular NoSQL database. It's also designed to be future-proof by being extensible to new interfaces and storage devices. 
"As the only company developing storage class memory, flash, and DRAM technologies, Micron is uniquely positioned to build a software stack that accelerates applications running in today's flash-based storage environments as well as storage class memory-based infrastructure of the future," said Derek Dicker, Micron's general manager of the Storage Business Unit.

The news isn't just that HSE is now open source. The storage engine also lowers latency, especially for large-scale data sets. It does this through intelligent data placement. 

How much better does it work? Micron claims HSE improves the throughput of particular storage applications by up to six times, reduces latency 11 times, and improves SSD endurance by seven times. It can also take advantage of multiple classes of media concurrently, such as flash and 3D XPoint technology. With Micron's own hardware, such as the Micron X100 NVMe SSD working with a set of four Micron 5210 QLC SSDs, Micron claims you'll see double the throughput with nearly four times the read latency improvement.

In addition, HSE delivers the following features: 

  • Besides its MongoDB Integration, it can also be integrated with other storage applications like NoSQL databases and object stores.
  • HSE is ideal when performance at scale matters, including very large data size, large key counts (billions), high operation concurrency (thousands), or deployment of multiple classes of media.
  • The platform is designed to be extensible to new interfaces and new storage devices, enabling use with a broad range of applications and solutions, beyond traditional databases, such as Internet of Things (IoT), 5G, artificial intelligence (AI), high-performance computing (HPC) and object storage.
  • HSE is capable of delivering additional performance for software-defined storage -- such as Red Hat Ceph Storage -- that enables cloud-native applications through containerized platforms like Red Hat OpenShift as well as tiered performance for file, block, and object storage protocols for multiple use cases.

"We see enormous potential in the technologies being introduced by Micron, especially as it takes an innovative approach in lowering the latency between compute, memory, and storage resources," said Stefanie Chiras, Red Hat's VP and general manager of Red Hat Enterprise Linux. "We look forward to working with Micron in the open-source community to further build out these innovations, ultimately bringing new choices to the storage world derived from open standards and concepts."

Related Stories:

Editorial standards