Browse Prior Art Database

Method and Apparatus to Make Storage Data More Resilient to Data Decay

IP.com Disclosure Number: IPCOM000245735D
Publication Date: 2016-Apr-04
Document File: 2 page(s) / 48K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method and apparatus to increase the resiliency of storage data against data decay, especially in media that uses an electric charge to discern between ONE and ZERO bits. The method decreases the overall data’s exposure to decay data by counting the set/unset bits in a given chunk of data that is about to be written to persistent storage and then accordingly decreasing the charged bits.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

Method and Apparatus to Make Storage Data More Resilient to Data Decay

The data storage industry is striving to save space in multiple ways. In the past, data compression and data deduplication were exercised only in backup and archive repositories because the methods caused ill effects on latency. The advent of dedicated compression hardware (HW) and sophisticated software (SW) algorithms have brought these very techniques to the primary storage devices. By reducing the data footprint, the importance and weight of each physical bit of data increases. Hence, every effort is made to protect the bits against corruption and loss.

One known phenomena in data storage is built around electrical charge and is the data decay event in which, over time, bits flip and change value. In addition, FLASH has some vulnerability.

The focus of this disclosure is the media that uses an electric charge to discern between ONE and ZERO bits (e.g., Secure Socket Drive (SSD), FLASH).

The novel contribution is a method to decrease the overall data's exposure to decay data by counting the set/unset bits in a given chunk of data that is about to be written to persistent storage. If the majority is ones (1s), then the uncharged bit is used for it and is indicated in the respective metadata for use/reconversion on read. If the majority is zeroes (0s), then the uncharged bit status is used. Because an uncharged bit cannot decay, this method limits exposure to decay of the overall data by decreasing the charged bits as appropriate. This approach is performed in storage cache, costing almost nothing.

This approach is especially beneficial for compressed files in which the number of set bits tends to outweigh the unset bits.

To implement the method in a preferred embodiment:


1. Chunk of data received is in storage device cache (normal storage process)

2. 'Resiliency Process' counts the number of set/unset bits for chunk of data from Step 1

A. If the majority of bits is 0s (uncharged), then the 'resiliency updated' bit in metadata is set to 'no'; data is

written to persistent storage (normal storage process)

B. If the majority of bits is 1s (charged), then the uncharged bit is used to store data
i. after conversion is complete, the 'resiliency updated' bit in metadata is set to 'yes'
ii. data is written to persistent storage (normal storage process)

3. As data is read from storage, if the 'resiliency updated' bit in metadata is set to 'yes', then the data is converted

1


Page 02 of 2

prior to sen...