Forbes

How Many IOPS Is Still Enough

By Tom Coughlin

In 2012 Coughlin Associates and Objective Analysis conducted a survey of digital storage end users from a number of different industries to find out about their storage needs. In 2016 we repeated the survey to see how things have changed since the original survey in 2012. We found that over the last 4 years requirements for IOPS, latency and capacity have increased significantly. The availability of higher performance as well as less expensive storage has increased consumer expectations.

Between 2012 and 2016 the survey asked over two hundred sixty respondents to share their high-speed storage capacity and performance needs, IOPS and maximum latency, along with the speed of the rest of their system and for the application type being run on that system. We compared the results from the 2012 report on How Many IOPS is Enough with the 2016 data in the 2016 report by the same name.

The majority of our survey respondents required fast storage for the following application types:

  • Database
  • On-Line Transaction Processing (OLTP)
  • Cloud and Storage Services
  • Scientific and Engineering Computing

The figure below shows the distribution of survey participant dominant applications from 2012 through 2016:

 

Figure from Coughlin Associates

 

Figure from Coughlin Associates

Those survey responses that changed between 2012 and 2016 generally represented increases in both speed and capacity and reductions in expected latency.

Most respondents (79%) required performance of over 1,000 but fewer than ten million IOPS over a broad range of capacities The figure below compares IOPS requirements from the 2012 survey with the post-2012 survey results.

 

Figure from Coughlin Associates

 

Figure from Coughlin Associates

There is a 54% difference in the mean of the two data sets, with the new survey results requiring a mean of 1,511,338 IOPS versus 978,749 for the older survey data. The median of the newer survey data is 37% higher than the median of the old survey data (39,540 versus 54,063 IOPS).

Latency requirements had a very solid peak with 32% requiring 10 milliseconds. The figure below compares 2012 report minimum latency requirements to those post-the 2012 report. The median latency of the newer survey data is 54% lower than the old survey data, 0.92ms versus 2.0ms.

 

Figure from Coughlin Associates

 

Figure from Coughlin Associates

In many cases there was a reasonable correlation between two of the surveyed parameters:

  • Capacity & IOPS requirements
  • Capacity & Maximum Latency
  • IOPS Requirements and Maximum Latency

The sample group was a respectable size, giving us a 6.0% margin of error at a 95% confidence level. We continue to collect survey data with a goal of increasing the respondent base and reducing the error margin.

The availability of higher capacity digital storage technology combined with higher performance storage technologies has increased customer expectations and needs. The How Many IOPS Is Enough survey shows these trends for IOPS, storage capacity and minimum latency for several important applications, including correlations between these different needs.

 

This article was written by Tom Coughlin from Forbes and was legally licensed through the NewsCred publisher network.