HDD’s, fans, air conditioning, power supplies as well as the adjacent racks are the primary sources of vibration in Data Centers. The “Vibration Penalty” impacts every Data Center but the most utilized and highest density Data Centers suffer the most, because vibration is cumulative

Vibration has not been entirely ignored. The IT industry has traditionally viewed episodic vibration like construction or seismic events as a system failure mode. Ongoing levels of vibration, which is not as well understood by the industry, degrades disk storage performance, availability, and efficiency.

Vibration is a significant factor that affects the performance and efficiency of storage. Vibration increases disk latency which reduces disk drive performance, application performance and increases costs. Managing vibration is an important factor in realizing full value from hardware investments.

Since all Data Centers vibrate constantly, there is a good chance you have a vibration problem affecting your disk storage systems. When vibration reduces your storage performance and efficiency it also increases your costs by leading you to invest in additional storage hardware to achieve your needed performance level or Service Level Agreement (SLA). Over-provisioned storage takes more space, and uses more energy. Our VMA 1000 Vibration Monitoring and Analytics product will help you determine the amount of lost performance and efficiency due to vibration and help estimate performance increase and business benefits that can result from managing vibration.

We recommend installing our diagnostic/monitoring system VMA-1000 which includes a set of vibration sensors to measure the existing levels of vibration in your working environment and establish its impact on exiting storage performance. The monitoring should be used at least for 14 days or a representative production cycle so that the vibration modulation is recorded, analyzed and correlated with storage performance. With positive data, analysis, and expected savings in hand, the vibration remediation technology can then be applied to new storage as it is installed.

No. We correlate vibration level with storage performance parameters (IOPS and throughput) in your current performance tool. This can be done thru open API in your tool. GPC offers Ganglia, an open source performance tool, upon request.

It all depends on the vibration level in your storage as well the type of storage/HDD you have and the application you run. We believe you will see improvements in performance of at least 15%. Our testing in both labs and customer situations suggests that much greater improvements (30% or more) are possible. The best way to determine how VMS will impact your Data Center is to try it for yourself.

Information has strategic value for business; faster access to information is both revenue and a savings opportunity. Better decisions drive business for example to: parse claims data for insights, find bottlenecks in the supply chain, look for patterns in patient data, simulate a new product, improve price bids for oil leases, and analyze web content.

SSD’s have a role in improving disk storage performance, and the technology works best with structured data in small sizes. Performance of applications running large unstructured data (Big Data) are generally gated by native disk drive speed which is most susceptible to vibration.

Application availability will improve as requested data will be provided at a consistent speed, avoiding failures caused by excessive or unexpected delays in processing the data. These delays cause failures often referred to as a ‘soft error’ or 'time outs' within the industry.

Big Data is unstructured data and the size of the file to be processed can be measured in Gigabytes, Terabytes, and even Petabytes. It is the fastest growing type of data, and has been estimated to account for 80% of all business data. Big Data includes Video, Images, Scientific Logging Data, Web Content, and email, as well as other types of data that are not cache friendly because they do not fit conveniently into databases.