Virtualization Technology News and Information
You Can't Flash Your Way Out of a VM Performance Problem

Contributed Article Written by Brian Morin, Global SVP of Marketing, Condusiv Technologies

It's very common for virtualized organizations to run up against an I/O ceiling. When the application requires higher performance than the company's storage infrastructure can deliver, IT administrators often think the solution is to "just keep adding flash." But simply throwing more expensive server and storage hardware at the problem without understanding what's causing the real malaise is a mistake. Using more spindles or flash simply masks the problem and wastes resources. The fact is - you can't flash your way out of a VM performance problem.

That's because flash isn't a cure for I/O inefficiencies. It merely processes inefficient I/O profiles more quickly than disk, leading some administrators to think the problem is solved, when in fact much of the investment in flash is being wasted on unnecessary cycles. Flash is a hardware-based treatment that fails to address the real symptoms of an inherent software problem, because I/O inefficiencies at the Windows OS and hypervisor layers end up robbing optimal performance. To truly cure and solve I/O inefficiencies once and for all, the root of I/O problems must be addressed. That's the only way that you can ensure that you are getting the most from the flash or spindles you already have in your organization before spending money on more that you don't need.

Since hardware can only process I/O, not optimize I/O, a better solution is to consider I/O reduction software that targets the problem of increasingly smaller, more fractured, and more random I/O that penalizes performance the most. By sequentializing the I/O stream from VMs and targeting the most problematic I/O to be serviced by server-side DRAM, virtualized organizations can decrease I/O to storage by 50% while increasing application performance from 50-300% on the hardware they already have.

Ripping Off the Band-Aid

As Storage Switzerland LLC notes, instead of rushing into simply adding more and more flash for more performance, data centers are much better served by first examining solutions that optimize I/O where it is created before investing in expensive new storage technologies. Many IT budgets can't afford this type of brute-force approach that simply throws more hardware at the problem, creating an overreliance on hardware solutions that squander much of an organization's investment in flash or spindles. I/O reduction software, on the other hand, can significantly boost storage performance while helping administrators save money by addressing the root cause of I/O inefficiencies.

Today's I/O reduction software non-disruptively solves application performance issues in virtualized environments since the software is transparent set-and-forget, and built from the ground-up to operate with nearly zero overhead, relying instead on otherwise idle, available resources. It can thus optimize writes by increasing I/O density and sequentializing writes in addition to optimizing reads by utilizing available server-side DRAM as the first caching tier in the infrastructure. While DRAM isn't capacity-intensive, it's multiple times faster than a dedicated PCI-e or SSD cache sitting below it. Since DRAM is so fast and algorithms at the VM layer are application-aware, as little as 4GB of available memory consistently cuts latency times in half and alleviates I/O to storage by targeting the I/O that dampens performance the most - small, random I/O.

Since I/O reduction software sits at the VM layer where I/O originates, sophisticated algorithms can be employed that make cache churn almost non-existent by collecting and analyzing I/O data across a wide range of data points including storage access, frequency, I/O priority, process priority, types of I/O, nature of I/O (sequential or random), and time between I/O.

By sequentializing I/O profiles from the VM and prioritizing the smallest, random I/O to be served from DRAM, I/O reduction software ensures that the most performance-inhibiting I/O can't degrade the infrastructure. This software approach:

  • Protects a company's investment in its existing hardware infrastructure
  • Fixes performance bottlenecks without disruption
  • Helps organizations reap the maximum benefit from any future storage system investment whether with SSD or HDD
  • Eliminates concern about carving out limited DRAM for caching purposes, as the software itself leverages whatever DRAM is available.

In short, organizations lose up to half their throughput from server to storage from I/O inefficiencies yet are generally led to believe the antidote for the system's malaise is more IOPS in the form of expensive flash solutions. Once organizations discover they can address inefficiencies with I/O reduction software that eliminates half of the I/O to storage for any given workload, they aren't nearly as IOPS dependent as they once assumed and are able to solve their toughest application performance challenges without adding new hardware while getting the most from the hardware investment they've already made. If you're still in doubt about whether you should add more flash hardware for performance issues or leverage I/O reduction software, keep in mind that I/O reduction software is available for no-risk evaluation so you can see the actual before/after benefits in your real-world environment before purchase commitment.


About the Author 

Brian Morin, Senior Vice President, Global Marketing, Condusiv Technologies, is responsible for the corporate marketing vision by driving demand and awareness worldwide. Efforts over the last year led to growing adoption of V-locity, which has quickly amassed over 1,000 new customers looking to accelerate their virtual environment with a 100% software approach.

Prior to Condusiv, Brian served in leadership positions at Nexsan that touched all aspects of marketing, from communications to demand generation, as well as product marketing and go-to-market strategies with the channel. Brian notably steered rebranding efforts and built the demand generation model from scratch as an early marketing automation adopter. Growth led to the successful acquisition by Imation.

With 15+ years of marketing expertise, Brian has spent recent years on the forefront of revenue marketing models that leverage automation for data-driven determinations. His earlier background has roots on the agency side as creative director, helping companies build brands and transition to online engagement.
Published Thursday, August 06, 2015 6:41 AM by David Marshall
Filed under: ,
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<August 2015>