April 06, 2022

Intent-based provisioning | Think and it shall be done

Scott McGarry
Hewlett Packard Enterprise Solutions Architect at Data#3 Limited

Intent is an interesting thing.

It’s a determination to do something, have something actioned or achieve a result, but it’s not an outcome. In the technology space, it can be easy to forget that intent isn’t foolproof, and, despite our best efforts, there is often a big difference between our intent and the final outcome.

One simple example is the world of intent data. A mysterious, black art of marketing that attempts to glean your intentions when you search the web, read certain articles or visit particular websites to give you more relevant information and offers. What is it you’re really trying to achieve? Sometimes the predicted intent is right and sometimes it’s wrong, but by using it, marketers can save enormous amounts of time getting campaigns to a point where they can then make a more informed decision about what to do next.

Harnessing the power of intent has now come to data storage with HPE’s intent-based provisioning. It promises to do a similar thing – saving storage managers enormous amounts of time by eliminating the guesswork when provisioning and managing storage across on-premises and cloud environments – and it works.

We’ve been putting intent-based provisioning through its paces, and it really is an innovation in storage management, bringing a true cloud operational experience across your storage infrastructure.

Let’s go back a step.

Despite all the advances in storage technology, managing it can still be an admin-heavy experience with the complexity of LUNs, RAID groups, HBAs, load balancing, thin and thick provisioning, resource planning, and more. Trying to work out exactly what performance and storage you have, and then matching it up with the application workload and SLA requirements can take hours – if not days.

With intent-based provisioning, all the underlying complexity is abstracted away from you. You describe your requirements in terms of application type, capacity and performance, and it does all the grunt work for you, providing a recommendation on where it should be provisioned. You can choose to accept, ignore or modify the recommendation before any provisioning takes place.

Behind the scenes, it’s leveraging real-time context across the fleet using AI and what-if simulations to predict how applications are likely to behave and evolve before coming up with recommendations.

It avoids the guesswork and frustrations around over-provisioning (to minimise the risk of running out of storage) and under-provisioning (to maximise the utilisation of storage assets with the hope that actual requirements are less than expected requirements).

As mentioned earlier, it’s an intent-based service so it still requires a level of human interaction to validate the suggestions and make decisions on the next actions. So, some of the claims around intent-based provisioning allowing people with no storage expertise to manage storage are a bit of an exaggeration. It does however demonstrate the value of analysis work that can be done by AI-supported processes and tools to save time, which is also demonstrated with another piece of Hewlett Packard Enterprise (HPE) technology – CloudPhysics.


In a similar vein to intent-based provisioning, CloudPhysics does the underlying analysis to help you right-size application and workload requirements ahead of migration to new infrastructure, regardless if it is on-premises or in the cloud. It provides an ideal configuration per workload for compute, memory and storage by extracting complete configuration and resource utilisation data across virtual machines, physical servers and storage arrays.

It can compare requirements between on-premises and public cloud and model your exact infrastructure as a virtual environment so you can see what impact particular decisions will have. The incredible part of CloudPhysics is that in just 15 minutes you can get initial insights towards optimisation options with nothing to install or maintain.

It’s not just a one-trick pony for upfront recommendations either. It can continuously monitor your infrastructure to provide ongoing data-driven insights to ensure everything stays optimised from edge to cloud.

Seeing is believing

With Data#3 being one of HPE’s largest Platinum Partners in Asia-Pacific, we can provide a complimentary CloudPhysics insights report based on your current infrastructure. It can be a real eye-opener to see the level of detail and value such a quick process can provide.

HPE’s direction in this space is very clear and we expect to see further developments as they continue to ramp up efforts to utilise AI to do the heavy operational-based lifting across all parts of your on-premises and cloud-based infrastructure.

To take advantage of this CloudPhysics offer, contact us today.