Around the Storage Block
1777182 Members
3180 Online
109065 Solutions
New Article
StorageExperts

Optimal workload placement using the AI-driven HPE InfoSight Resource Planner

Learn how HPE InfoSight Resource Planner for HPE Alletra 6000 and HPE Nimble Storage can be used for planning hardware platform upgrades based on your workloads. Minimize risk of downtime caused by overuse of resources, and maximize hardware utilization to reduce costs.

HPE-InfoSight-Resource-Planner-Blog1_shutterstock_1727577790.pngIT administrators must continuously plan for IT hardware specifications to meet workload needs, whether it be for the addition of brand-new application deployments, or simply planning for growth.

They must determine if their existing storage hardware can service additional workloads, what the contribution of each workload is to storage hardware utilization, and what additional storage hardware needs to be purchased to accommodate new workloads, and so on. Currently, there is not much assistance available to IT admins to complete this activity. Many times they end up manually creating spreadsheets to estimate resource usage on storage hardware platforms, making this a cumbersome and time-consuming activity.

The newly announced HPE InfoSight Resource Planner was designed to help IT administrators efficiently plan hardware needs for their workloads, while minimizing the risk of downtime caused by overuse of resources, and maximizing hardware utilization to reduce costs. This new version of the Resource Planner is an upgraded and improved version of the Resource Planner tool utilized in our Labs environment that was released 2 years ago. The HPE InfoSight Resource Planner is now available worldwide, so customers outside of the US can also benefit from its features.

The IT administrator can request sizing of new workloads or migrate and scale existing workloads from one unified dashboard, to review historical performance of workloads. Sizing is based on machine learning models in the backend that predict IO, capacity, and cache headroom for each hardware model. The Resource Planner provides a sorted list of recommendations of preferred purchased and new storage arrays that the workload could run on. Three main workflows place new workloads on existing and new systems, scale and migrate workloads onto both new and existing systems.

HPEInfoSightResourcePlanner-3MainWorkflows.png

New workloads can be defined based on data gathered from our global installed base. Users can also override the default HPE InfoSight model based on additional business context. For migrating and scaling existing workloads, admins can specify the time range of the workload to model. The Resource Planner predicts resource usage on storage arrays and attributes the resource usage to specific workloads. For example, IO/Capacity/Cache usage on the storage array will be attributed to specific workloads that are simulated on the storage array.

HPEInfoSightResourcePlannerAnalysisResults.png

Recommendations can be further customized per the IT admin’s additional business context, and they can filter recommendations based on remaining storage resource headroom. For example, the user could be risk averse and choose to keep the resource headroom buffer at 50%, in the event that there is a spike in usage. Or the user could have medium risk tolerance, and will lower cost by reducing the headroom of the remaining buffer to a smaller amount. Another case could be that some workloads are not open to migration, because they need a lot of lead time to accommodate purchase of new hardware; therefore there is a need for additional headroom.

HPEInfoSightResourcePlanner-Filter Results.png

Sizing storage is a challenge because bottlenecks can be at various layers of the architecture. HPE InfoSight provides AI-driven assistance that takes into account the complexity for each application, while supporting a broad set of commonly used workloads. Workloads are modelled on specific characteristics for applications like VDI, Oracle, and SQL Server, among others, as well as several other axes like dataset size, performance, and data protection. Simulating impact on storage array resources – like those related to IO, cache, and capacity – are based on past history. Analytics are based on data gathered from customer environments over a period of several years, and are refined and improved over time.

The HPE Resource Planner also powers the Intent Based Provisioning tool in the Data Services Cloud Console.

Watch the demo video for a complete overview of its functionality.

Try out the HPE InfoSight Resource Planner for yourself! Log onto the HPE InfoSight portal and navigate to the HPE Alletra 6000, HPE Nimble Storage menu. Then, expand the new planning section and choose Resource Planner – and see for yourself how simple it is to plan hardware platform upgrades based on your workloads.

HPE InfoSight Resource Planner Menu.png

 

 

For more information about other services powered by HPE InfoSight Resource Planner, please check out this blog: Reimagine storage provisioning: Intent-based provisioning with Data Services Cloud Console.

Anshul headshot medium.jpgAnshul Madan is passionate about reflecting the voice of the customer, and building valuable products that delight those users. He is a Product Manager for HPE InfoSight where he is responsible for workload insights and other forward-looking AIOps features. Anshul enjoys reading, yoga, spending time outdoors, and playing sports.

You can connect with Anshul on LinkedIn!

 

 


Storage Experts
Hewlett Packard Enterprise

twitter.com/HPE_Storage
linkedin.com/showcase/hpestorage/
hpe.com/storage

0 Kudos
About the Author

StorageExperts

Our team of Hewlett Packard Enterprise storage experts helps you to dive deep into relevant infrastructure topics.