What are All-Flash Arrays? | Tintri

0 0

What are All-Flash Arrays?

An all-flash array (AFA) is storage consisting of flash memory within solid state drives (SSDs) with integrated circuit assemblies, as opposed to hard disk drives (HDDs) storing data on spinning platters. All-flash arrays come in many sizes and types with options for networking and infrastructure, but they always retrieve data more reliably and efficiently than HDDs because there are no moving parts. Whether they’re HDD or AFA, however, the right software can help all-flash arrays and hard disk drives operate on the level of virtual machines (VMs) and containers: perfect for enterprise storage needs.

Why enterprise is moving toward all-flash

All-flash arrays (AFAs) read, write, and execute computations with far greater speed than HDDs. But other factors make AFAs a better choice for enterprise:

  • Dramatically reduced latency events. HDD storage can take minutes to retrieve data or execute commands, while AFAs take seconds—or less. What’s more, when paired with the right enterprise software, DevOps teams can find the root of the latency at the most granular level: virtual machines (VMs) and containers. That means quicker troubleshooting, faster repairs, and virtually no downtime for your venture.
  • AFAs lower cost within the data center. They save between 50–80 percent of current spending on HDD storage (depending on your infrastructure, whether it involves only physical servers, or if you’re operating in a hybrid environment, etc.) That's because AFAs generate much less heat—a sign of high-energy use in and of itself—and do not require the expensive cooling HDDs need to keep from literally burning up.
  • AFAs take up so much less space on racks within data centers—about one-fourth the size—so you can fit more storage, or opt to stay small until it’s time to scale up.

The heart of AFAs: nonvolatile NAND flash memory

NAND stands for “negative-AND,” referring to the logical gates that NAND flash arrays’ floating-gate transistors resemble. NAND is the preferred method for flash storage today because it’s more robust and reduces the time it takes to read, write, or erase data, code and software.

The nonvolatile characteristic offered by NAND cannot be found in earlier arrangements of floating-gate transistors for flash storage, such as NOR. As nonvolatile storage (NVS), NAND is critical to enterprise flash storage arrays. Nonvolatile means that your AFA enterprise will not lose any data, software, applications, or any other data in the event the power goes out. Every VM and container on your AFA are safe with nonvolatile NAND AFAs.

Advantages and disadvantages of AFA versus magnetic disks on SSDs

Some of the heaviest hitters in the consumer space, including Facebook, Amazon, and Dropbox have switched—or are in transition toward—AFA storage. But for the small enterprise, the perceived cost of AFA can thwart any CFO’s desire to migrate to AFA.

Killing the rumor: flash storage prices have dropped dramatically

Not that long ago, the argument against transitioning from legacy storage to AFA was that all-flash was astronomically more expensive. Today, that’s simply not true.

Tintri All-Flash Array offers what your enterprise needs

With the Tintri All-Flash Array, you’ll get the agility of public cloud in your data center. Conventional AFAs just aren’t suited to work with cloud-native applications. With Tintri CONNECT architecture, application performance, IT automation, and predictive analytics for scalability are guaranteed.

If you’re still working with legacy storage and you’ve tried to reconfigure the LUNs and volumes within your HDDs to support a cloud environment, it’s time to make a change. Have a look at the agility, flexibility and scalability Tintri application-level all-flash storage will bring to your enterprise cloud.