Browse Prior Art Database

Guaranteed Higher Performance for Critical Applications in Solid State Drives

IP.com Disclosure Number: IPCOM000240806D
Publication Date: 2015-Mar-04
Document File: 5 page(s) / 215K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a way to guarantee higher performance for critical application in solid state drives (SSD) by reserving a designated space, away from background operations, and using that for faster turnaround time when the application demands it, based on priority

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 5

Guaranteed Higher Performance for Critical Applications in Solid State Drives

Solid State drive (SSD) performance varies with respect to current state of the data written in the drive. Disk drive (solid or hard drive) performance is often related with multiple attributes, and key to an integral system level performance will be the I/O operations per second and the data transfer rate (MBPS). Factors that aid in affecting the data throughput are read/write percentage, transfer size of the data, and sequential/random addressing.

    In an enterprise storage scenario, one really can't dictate the read/write and sequential/random values since that depends on the kind of workload that the user is running. Application and the operating systems, to some extent, are aware of the type of the data written into or read from a disk drive and their priority. Can higher priority applications demand a guaranteed performance when data is written into and read from?

    The novel idea of guaranteeing peak performance in SSD at any point in time during the operation of the system can be achieved by:

Reserving a designated space for faster access, away from the reach of background operations


Opening up that space for faster turnaround time while writing data


Mask that designated area for easier access to read and write, bypass all the other algorithmic check that is done prior to data fetching and writing


Synching the temporary buffer space internal to the SSD along with the priority also helps in better data movement


Always maintaining a pool of fresh PBA intact from background activities help in serving multiple high-priority applications


Regular handshake between the host and drive on the availability of such reserved areas


Varying the size dynamically based on host request by employing a standard software piece running in the host

Space Reservation and Faster Access

    An empty space of defined size allocated by the SSD software in agreement with the host and not touching the space until the host commands the SSD to. (see Figure 1 below). Varying sizes or pockets of reserved area as defined by the host, as host knows what kind of priority application it's dealing with. Denying accesses of all the background operations to that particular reserved space, so that it can be available at any point in time. Maintaining that space as fresh out-of-the-box equivalence, so that performance is at its maximum when data is written or read from. Simple address...