In order to ensure cloud servers are performing at their maximum capacity, public cloud service providers need their storage to act intelligently and maintain security while keeping costs low. Using the Pliops Storage Processor, as displayed during June’s Cloud Field Day event, IT departments can achieve the goals of performant cloud server stacks — regardless of whether they’re disaggregated or dedicated.
Disaggregation versus Dedication in the Cloud
Traditionally, cloud server deployments have relied on disaggregation to ensure that they can draw upon required resources as necessary while remaining secure. Pulling from pools of compute and storage, these servers can be used dynamically, but when used to run dedicated applications, often end up carrying the potential to waste resources that are left unallocated, and therefore, unused.
An alternative method for producing cloud servers is to instead design around the application, building everything needed to run the app at max performance and ensuring the server does only that. In this scenario, oftentimes, a more intelligent approach to storage is needed. “Dumb” storage options can consume valuable compute resources when writing and allocating storage, so having storage with processing power built-in can save both time and resources, maximizing output.
Pliops Storage Processor is Built for the Cloud
At the end of the day, regardless of whether you opt for disaggregated or dedicated cloud server infrastructure, Pliops and their Storage Processors are ideal for ensuring the utmost performance at the storage level. At their core, Pliops Storage Processors are designed to move the needle in 4 areas that are critical to creating an effective cloud server: reliability, performance, capacity, and efficiency.
A Pliops Storage Processor can be plugged into any PCIe slot in a server and uses its built-in processing power to handle storage-related compute processes, reducing the overall load on the server as a whole while ensuring that tasks are completed efficiently. It integrates well with many of today’s top database providers as well, including MongoDB, MariaDB, MySQL, and others.
Building Pliops into the Cloud
At June’s Cloud Field Day event, Tony Afshary, Senior Director of Product for Line Management at Pliops, displayed the Storage Processor and outlined its capabilities, especially when used in disaggregated infrastructure. During his presentation, Afshary outlined the four goals of disaggregated cloud infrastructure that the Pliops Storage Processor supports:
- Isolation: Separates resources to minimize attack vectors
- Security: Minimize uncontrolled East/West traffic to prevent attacks
- Scale: Fits into nearly any server, so can be built en masse
- Cost: Keeps server spend low by maximizing output and efficiency
Afshary then went on to showcase how Pliops is being used by public cloud providers to accelerate their deployments at massive scales. With Pliops added to server clusters, availability and uptime are kept as high as possible since Pliops allows cloud providers to stack multiple SSDs in groups, rather than leveraging individual drives.
Ultimately, thanks to these larger, more reliable clusters, Pliops customers have experienced up to a 45% expansion rate across their server stacks with zero downtime and a reduction of cost of up to 77%. At the massive scale of public cloud deployments, such a capability essentially doubles the output of these servers at a fraction of the cost, allowing for the creation of more servers performing at the same great rate.
Zach’s Reaction
When it comes to maximizing performance and uptime in the cloud, Pliops ensures that data is stored smartly and efficiently while keeping costs low. For such large scales that public cloud offerings usually require, adding Pliops to the stack seems like a no brainer.
Watch Pliops’ follow-up presentation from their Cloud Field Day appearance to get an even deeper technical dive into their product, and be sure to visit their website for more information.
