This article is more than 1 year old
Embarking on your digital transformation journey? Make sure you fix your data first
Increasing data flow, with HPE GreenLake for storage
Paid Feature Data is the lifeblood of the modern organisation, essential for fuelling those higher level functions like data driven analytics and AI. And, just as with humans, if the systems that maintain, refresh and circulate that lifeblood are neglected, the outcomes are…well, not great.
Consider recent research by ESG for HPE, which shows that "data first" organisations – those with the most mature data strategies - are typically 20 times more likely to beat their competitors to market, and more than 11 times more likely to beat their revenue goals by 10 percent or more. Data first organisations also have much happier data analysts and scientists.
So what defines a "data first" organisation? They're the ones who have confronted the complexity of data management – freeing them up to actually do something more useful with their data. Over 93 percent of respondents to ESG's research see complexity holding up digital transformation, with 59 percent saying complexity is "trending in the wrong direction". The result is that three quarters of organisations are unable to keep pace with business requirements.
Typical warning signs for data complexity include running multiple storage platforms, each with their own management tools, which becomes a progressively tougher problem at scale. Likewise, companies find they need myriad complex provisioning tools, and face ongoing problems in meeting service level agreements for their applications. And when it comes to addressing these problems, they have to navigate glacial capex procurement processes and - at a time of supply chain snarl ups - drawn out delivery schedules.
This firefighting might sound like business as usual to many tech pros. But the fact is, says HPE Storage's Worldwide Product Marketing Manager, Simon Watkins, "they're spending too much time managing the infrastructure, as opposed to really leveraging the data."
The public cloud at first glance might appear to be the perfect solution. Cloud promises all the infrastructure you want or need, with straightforward, predictable pricing. That same ESG research points to how IT leaders see the problem playing out, with more than nine in ten agreeing that "mature cloud operations on-premises is the single most important step to eliminate complexity".
It is certainly the case that lines of business (LOB) staff and app owners as well as IT leaders and admins appreciate the simplicity of managing hardware in the cloud. Updates and patching are no longer their concern, provisioning is theoretically much more straightforward, and capacity is unconstrained. And, once the initial transition is made, procurement becomes much less problematic. No big ticket capex cycles, no more supply chain delays, and no more over-provisioning.
But experience has taught many companies that things are not quite that simple. Cloud pricing might appear straightforward. But platform providers' byzantine pricing changes regularly, while the very ease of consumption can undermine procurement discipline. When hidden extras are worked in, that predictable pricing can translate into monthly sticker shock, especially when hefty egress charges are levied.
Keep the data flowing
Equally importantly, many companies are not in a position to move all, or even some, of their data to the cloud. The reasons range from concerns about data gravity and latency to exposing IP, to unavoidable data sovereignty or privacy restrictions. So, there has been no shortage of offerings aiming to reproduce the cloud experience for on-prem, or hybrid environments. To offer storage "as a service" so to speak. But there is another catch.
Often, the focus is on the consumption part of the equation. On a purely hardware basis, this typically includes the ability to unlock extra capacity when needed, utilising built-in buffers. Beyond the box, these solutions are offered on a consumption-based pricing model, effectively making them opex rather than capex buys. This goes some way to providing more predictable costs, as well as providing other financial benefits – and partially addresses the supply chain conundrum.
But looking back at ESG's research it's worth remembering that most IT and business professionals hanker for the experience of "mature cloud operations". That means much more than being able to buy on-prem capacity more easily.
"What we're hearing loud and clear is that customers love the operational simplicity of the public cloud," says Watkins. But how they pay for it is just part of the appeal. What they really want to bring on-prem is "the automation, the flexibility, the agility."
In other words the novelty of being able to unlock additional capacity quickly wears off if IT admins must still use multiple management tools to provision and manage different classes of storage.
Likewise, tech managers need to know that this "storage as a service" can run any workload without compromise, whether that workload is a legacy application that can't be lifted and shifted to the cloud, or a modern application that needs to access resources on-prem or in the cloud. Having infrastructure on tap is not a great benefit if it still takes a day and a half to work out which is the best storage for a new workload.
HPE looks to satisfy this far broader requirement for cloud-like experiences with the HPE GreenLake platform which offers a wide portfolio of on-premises cloud services - spanning compute, networking and storage as well as database, HPC, containers, AI and VDI and third-party applications. HPE GreenLake for storage features a portfolio of on-premises cloud data services, offered in mission critical, business critical or general purpose options.
With HPE GreenLake for storage, data infrastructure is delivered on-premises, and charged on a consumption basis, which is calculated on the storage actually used after the data has been deduped and compressed. This satisfies those dual needs for control and a more flexible and predictable procurement methodology.
Abstracted, not distracted
Under HPE GreenLake for storage, the management of the storage infrastructure is abstracted into the cloud. This means that admins get a single console to manage their entire fleet of storage infrastructure, whether distributed across their real estate or in co-lo spaces. This allows them to simply manage, or offload, day-to-day provisioning and management activities such as patching, and set policies and SLAs centrally right across their estate.
Moreover, this is underpinned by AI, in the form of HPE InfoSight. As Watkins explains, this AIOps software for infrastructure offers full stack visibility, right up to the application layer. Maintenance is less a matter of ongoing firefighting, more a question of predictive support automation and continuous optimisation.
The platform also offers policy management and intent-based provisioning, so that deciding where an application should be deployed to meet required service levels is no longer a combination of exhaustive analysis, combined with an element of hope and pray. Rather, admins can specify the workload, the capacity and performance needed and the key policies, and the service will automatically recommend the best-suited system across a fleet to optimise SLAs.
This is particularly important as tech demands are increasingly driven by line of business imperatives, says Watkins. "They're the ones that really like the agility of the cloud when it comes to spinning up your infrastructure and they're the ones who get frustrated with the slowness that they get on-prem. They have to wait too long to have new infrastructure, or get new apps provisioned."
With HPE GreenLake for storage, he says, this bottleneck is removed. "Whether it's a modern app or traditional one, with capabilities like intent-based provisioning, you can cut storage provisioning times from days or weeks to minutes, and automatically ensure that workloads are provisioned on the right platform - with no storage expertise required."
The sum of all this? One easy metric to grasp is increased availability, to the extent that HPE GreenLake for storage offers 100 percent availability guarantee on its mission critical data service with HPE Alletra 9000 systems. More broadly, Forrester Research suggests that HPE GreenLake for storage can help slash time to deliver projects by 75 percent, a massive advance in a world where customer satisfaction and speed to market is everything. Likewise, Forrester expects 30 to 40 percent reduction in total cost of ownership, due to the elimination of over-provisioning.
But arguably the biggest benefit comes with the time savings that a centralised, cloud-like management of your entire storage infrastructure delivers – and potentially, of the other infrastructure and services accessible through HPE GreenLake To take one example, cited by Watkins, "LOB and database admins are happier because they can now move faster...e.g. they can self-provision storage infrastructure without the need for domain expertise."
Time saved can be devoted to other projects that will deliver on the potential of that data and unleash the digital transformation and innovation that every organisation covets. And if it all means that storage pros' own blood pressure evens out a little, well surely that's a good thing too.