Flash Aware Applications

Nisha Talagala (Fusion-io)
Sponsored Sessions
Location: E142
Average rating: ***..
(3.50, 2 ratings)

In traditional computing architectures, applications and file systems have been to limited accessing storage through basic block I/O operations. These operations, primarily read() and write(), have been handled by the block storage stack. Until the introduction of solid state storage technologies such as NAND flash, there has not been a compelling reason to change this.
However, non-volatile memory technologies like NAND flash have unique properties that require comprehensive management of the underlying memory media to provide good performance and data integrity. This kind of management is provided by a Flash Translation Layer (FTL) that virtualizes the physical blocks of the underlying device, providing a consistent Logical Block Address (LBA) range that is backward compatible with existing block device interfaces while also providing new capabilities.
Rather than treating flash devices simply as ‘fast disks’, it is now time to move beyond traditional block I/O.
We present the motivation for and describe examples of open source flash aware applications. Instead of using traditional layers of abstraction between applications and block storage devices, flash aware applications use direct programmatic access to flash to bypass software overheads and make use of new FTL based primitives.

This talk will describe the following:

  • Flash-aware open-source database applications
  • A Key Value interface to flash (enabling NoSQL databases to become more flash aware)
  • Making the Linux virtual memory swap subsystem flash-aware
  • How these applications make use of new flash based primitives such as Vectored Atomic Writes, Persistent TRIM, EXISTS and Sparse Access
  • Performance, endurance, memory consumption, and overall efficiency improvements possible via this approach
  • An overview of work in standards bodies and operating systems to support these capabilities

Nisha Talagala

Fusion-io

Nisha Talagala is Lead Architect at Fusion-io, where she works on innovation in non volatile memory technologies and applications. Nisha has more than 10 years of expertise in software development, distributed systems, storage and I/O solutions, and non-volatile memory. She has worked as technology lead for server flash at Intel – where she led server platform non volatile memory technology development, storage-memory convergence, and partnerships. Prior to Intel, Nisha was the CTO of Gear6, where she designed and built clustered computing caches for high performance I/O environments. Nisha also served at Sun Microsystems, where she developed storage and I/O solutions and worked on file systems. Nisha earned her PhD at UC Berkeley where she did research on clusters and distributed storage. Nisha holds more than 30 patents in distributed systems, networking, storage, performance and non-volatile memory

Sponsors

Sponsorship Opportunities

For information on exhibition and sponsorship opportunities at the conference, contact Sharon Cordesse at (707) 827-7065 or scordesse@oreilly.com.

Contact Us

View a complete list of OSCON contacts