“Data is the new source code,” as Ken Grohe, President & Chief Revenue Officer (CRO) at WekaIO, puts it in his trademark succinct manner, a sentiment he shared with me during a recent conversation on my podcast.
The true potential of your business is often hidden in the genetic code of your business’ ability to harness data. The question really is whether your current infrastructure is up to the challenge of processing data at ever-increasing efficiencies to drive business insights that can power everything from innovations to market opportunities, improve Return-on-Investment (ROI) and shorten Go-to-Market (GTM) cycles.
For many of us, unfortunately, personal experience with high volume data computing automatically harks back to lingering muscle memories of frustrating latency issues, version control and overall poor computational performance. But the good news is that handling limitless data at higher capacities does not automatically equate to poor performance.
Weka File System
The answer to high capacity performance computing for modern enterprise workloads under extreme conditions is pretty simple – a robust file system solution. A modern parallel file system is essential to get a ROI out of any data-intensive workloads at massive scales, such as artificial intelligence (AI) initiatives. Weka’s “Limitless Data Platform”, built on the Weka File System (WekaFS) is a software-defined architecture which allows organisations to remove hard limits on capacity scaling, and drive industry-leading performance and efficiencies in hyper scale storage.
In my mind, data is fast-turning into an asset to be placed on balance sheets, and should have key performance indicators (KPIs) placed on it, and be managed like any other mission critical asset (such as, human resource, plant & equipment, cash & liquid “assets” etc.). To that point, a data-performance enabler like a software-defined file system can provide you with the competitive edge you need to seamlessly consolidate all your workloads across data silos and drive high performance.
As one of the leading solutions in this market segment, the Weka File System (WekaFS) addresses this by leveraging best in networking technologies like NVMe-oF, NVIDIA Mellanox InfiniBand, 100Gb Ethernet, as well as advanced computing technologies like GPU acceleration.
According to Ken, the WekaIO Limitless Data Platform is an agile, flexible, safe, secure solution and can deliver up to 75% reduction in storage cost per unit, as seen in the case study with Genomics England’s massive gene sequencing project to sequence 5 million genomes by 2023.
With all of that in mind, I thought I would circle back and highlight for you, my top three takeaways from my conversation with Ken on our recent podcast and how they relate to these very issues from a business, technology and operational point of view.
AI and Big Data initiatives are going to be the drivers for big business results in the next decade. WekaIO enables successful AI and Big Data outcomes at a fraction of storage and maintenance costs. For instance, Ken mentioned that it takes only half the staff to run WekaIO than it does to run any NAS, SAN or cloud-based solutions. From my perspective, however, WekaIO’s enterprise-ready and hybrid-cloud features enable it to transcend AI and Big Data use cases, and lends itself to back up and DR strategies, IT agility, and lower CapEx.
There is naturally an expectation that WekaFS file system is a pathway to high performance computing on the public cloud, but the question that arises is “how do companies go about transitioning from existing NAS or SAN solutions and what benefits can they expect to see in the short and long-term”?.
Ken and I discussed into this during our recently podcast conversation, but it struck me after we had recorded the show that we would certainly get enquiries from our listeners for more information, and indeed we did – so I reached out to Ken and asked him to expand on this, and here is what he had to say:
“We have found that legacy systems tend to be inefficient and ineffective in supporting today’s demanding data-intensive applications, often forcing customers to make compromises on simplicity, speed, or scale.
Legacy storage technologies tended to be purpose-built to solve different problems, but when implemented as a system, produced islands, or silos of storage. This, in turn rendered the data invisible to the applications which needed it, resulting in a big problem for data centers today.
To address this, Weka’s Limitless Data Platform has been designed to offer a cost-efficient storage system, that combines simplicity, speed, and scale – shattering the storage limitations which continues to constrain enterprises from achieving better business outcomes.”, Ken Grohe
Perhaps the greatest technical differentiator I can see with the WekaIO solution, beyond obvious advantages in scalability and flexibility is its ability to ‘snapshot’ data instantly while allowing for updates, patches, testing or training to run without impacting the active ‘production’ copy of the file system.
Being a skeptic, I couldn’t help but wonder what would happen if the system came under stress, say from the likes of either a cyberattack or extreme high I/O workloads. So, naturally, I again reached out to Ken and asked him if he could give us further insights beyond what we discussed on the podcast, around the key technical advantages the Weka file system (FS) offered in this context, and his reply was enlightening:
“The WekaIO Limitless Data Platform does not rely on traditional back-up and/or disaster recovery scenarios which can often be compromised by cyberattacks or extreme high I/O workloads, as you mentioned Dez. We work with them, but incrementally we embrace a ‘snap to object’ or ‘snap to cloud’ approach, which bypasses those antiquated dependencies, and adds durability.
As a result the desired workflow can be achieved instantaneously by leveraging cloud, object, or on-premise NVME-tier storage, in scenarios such as cyberattacks or performance / unavailability, within the safety and dependability of a unified namespace.”, Ken Grohe
One of the biggest purported benefits of WekaIO solution, has been its ability to seamlessly integrate between on-premise, third party and cloud storage locations, from potentially small beginnings of tens of terabytes, to multi-petabyte scale, and to do so more cost effectively and infrastructure & human resource efficiently.
Given the difference in speeds inherent to each architecture type, I wondered if there might be potential performance bottlenecks or security issues which needed to be considered either from a design, implementation or operational points of view, as organisations attempted to assimilate decades old ‘frozen’ data on flash and disk with highly dynamic cloud storage.
This could be particularly relevant for data management challenges of large enterprises and government departments or agencies, or even financial institutions. Again I found myself realising the best person to answer this was Ken so once more I put the question to him, and his insights were once again on point:
“Dez to your point, a recent study from Forrester revealed that over 72% of all corporate data created is trapped in data islands. It is often marooned in separate SANs, NAS file shares, object storage, or even public or private / hybrid cloud platforms, in turn creating new islands, mostly never used.
Additionally, we have found that incompatible tools and manual copy procedures often make data sharing and data migration nearly impossible. We believe WekaIO’s enterprise-grade security, both in-flight and at-rest, provide the ideal data platform across any protocol and diverse workloads, while adhering to modern compliance policies.”, Ken Grohe
In a recent study IDC estimated that approx. 1.2 zettabytes (1.2 trillion gigabytes) of new data was created in the year 2010, up from 0.8 zettabytes the year before, and it has been estimated that the amount of the newly created data in 2020 was in the order of 44X to reach 35 zettabytes (35 trillion gigabytes).
With data usage growing at such an eye-watering fast pace, and massive scale, there can be little doubt that it is now time for enterprises of all sizes to start thinking about smart and efficient storage that can serve their data needs not just now, but help them capitalise on their data accrual five years down the line.
If you haven’t already tuned into the conversation I had with Ken recently on my podcast, please do via the link below, and I look forward to continuing this conversation with Ken and the team at WekaIO in further discussions soon.
- My podcast conversation with Ken Grohe: http://bit.ly/conversations-with-dez-podcast-featuring-ken-grohe