Comments
Why Flash Storage Benchmark Testing Is Not Hype
Newest First  |  Oldest First  |  Threaded View
storman2
50%
50%
storman2,
User Rank: Apprentice
10/7/2013 | 7:46:30 PM
re: Why Flash Storage Benchmark Testing Is Not Hype
The key is truly understanding the performance requirements of the workloads themselves and how they interact with the storage infrastructure. As George points out, relying on IOPS is a poor indicator of performance, particularly for file-based environments, which usually have very heavy metadata performance constraints. IOPS numbers don't take metadata performance into consideration. We at SwiftTest (ww.swifttest.com) are directly addressing this with workload modeling and performance validation solutions that are a direct reflection of your actual workloads. With this insight into your production applications, you can determine which storage systems and which configurations are truly best for your environment before putting them into production.


Cartoon
White Papers
Register for Network Computing Newsletters
Current Issue
2014 State of Unified Communications
2014 State of Unified Communications
If you thought consumerization killed UC, think again: 70% of our 488 respondents have or plan to put systems in place. Of those, 34% will roll UC out to 76% or more of their user base. And there’s some good news for UCaaS providers.
Video
Slideshows
Twitter Feed