vRealize AI Cloud: New BETA announcements

At VMworld 2020, we introduced the availability of VMware vRealize AI Cloud, a self-tuning service that uses reinforcement learning to continuously adapt to the changing needs of your VMware SDDC and dynamic workloads for performance optimization.

Since then, we’ve included in vRealize Operations Cloud with vRealize Cloud Universal subscription and also made it available in its very own stand-alone UI.

Now at VMworld 2021, we are excited to announce new BETA features for vRealize AI Cloud to support vRealize Automation Cloud and network optimization that is powered by vRealize Network Insight Cloud!

Let’s jump right into it! First, we’ve expanded on the vRealize AI Cloud self-tuning service to now offer the Storage Policy Genie recommendations engine to improve the performance or capacity of a cluster or virtual machine, based on changes to the Storage Policy-Based Management, and other configuration changes such as deduplication and compression. The Storage Policy Genie engine analyzes and predicts performance, capacity, and cost of change to trigger recommendations based on a net benefit score.

Coming soon as BETA, the Storage Policy Genie will make recommendations for clusters and virtual machines with existing workloads deployed with capabilities to:

Recommend changes to cluster and VM level default policy
Predict performance improvements, capacity impact and cost of change
Present multiple alternatives with differing trade-offs and explanations

vRealize Automation Cloud Optimization

Smart Initial Deployment

The complexity of modern applications and their supporting infrastructure is so high and increasing so quickly that soon only other applications will be able to understand them. Workload placement is typically based on manual triage or static policies not necessarily aligning with business goal.

Smart Initial Deployment learns patterns for better place workloads through ML modeling application level KPIs (cost, capacity, and licensing) for best performance. It identifies impacts on existing workloads to preserve performance and is based on goal-oriented placement.
Outcome:
- Smart placement based on KPIs leading to better performance and alignment with business value

Smart Initial Sizing

Virtual machines and workloads are usually deployed with max recommended resource allocations that typically require readjustment and monitoring. vRealize AI Cloud Smart Initial Sizing aims to maximize available capacity and avoid over provisioning to address no negotiation/less negotiation to right-size workloads through specialized modeling optimization algorithms.

vRealize AI Cloud Smart Initial Sizing addresses Day 1 challenges vs Day 2 challenges with right-sizing in vRealize Operations which helps minimizes monitoring of virtual machine utilization after deployment. It maximizes available capacity to avoid over provisioning and reduces reclaiming of resources with no hot-remove without shutdown/restart of VMs.
Outcome:
- Maximize available capacity and decrease costs with Rightsized workloads at deployment

Network Optimization by vRealize Network Insight Cloud

“The amount of data traversing over networks is so astronomical that it has become impossible for humans to process, analyze and act on it” – Extreme Networks on AI/ML

vRealize AI Cloud is expanding support to networking layer to dynamically finetune networking tunables for reduced latency, increased throughput, reduced re-transmits times and packet drops. For this upcoming tech-preview, support will be applicable to the virtual distributed switch (vDS)

Outcome:
- With features across storage, compute and networking domain optimization coming together, vRealize AI is one step closer to its vision of self-driving datacenter.

For more information

To learn more on vRealize AI Cloud, and other vRealize products, I invite you to check out our new VMware Pathfinder site that that covers detailed usage for: