Cloud Agility and Autonomous Operations: How AIOps Works from Edge to Cloud


Transportation systems are significantly safer and less likely to fail than they were just a few years ago – and the AI ​​behind that change is driving this trend. The technology has moved from basic controls to advanced features such as adaptive cruise control, anti-collision braking and sophisticated navigation which has dramatically improved road safety and user experience. Cars today are equipped with sensors and cameras that feed the data to an on-board processor that leverages the cloud for software updates and feature upgrades, keeping vehicles out of repair shops and on. the road safely.

Modern cars are essentially mobile computers or edge systems that house a combination of sensors and distributed processing capabilities located at the edge of the network where they collect information locally. Relevant data is transferred to the cloud for analysis and to train machine learning (ML) models. Refined models are then deployed in the vehicle to enable actions based on real-time sensor data and integrated AI, a self-training method that will one day enable fully autonomous vehicles, using a pipeline of End-to-end AI, from the edge to the cloud.

In IT, it’s time for infrastructure providers to deploy the same strategy and unleash the operational agility of the cloud with fully autonomous operations. As I pointed out in a previous article, Edge-to-Cloud AI is an essential component of an AI-driven infrastructure. I briefly alluded to an architecture that perfects this same process: collecting data from storage infrastructure devices, using that data to train ML models in the cloud, and deploying the final ML model to the devices. . It is essentially an enterprise storage system with integrated AI and machine learning in the cloud, forming a self-contained AIOps framework that spans from the edge to the cloud.

How does AIOps work from the edge to the cloud?

Traditional ITOps platforms are either confined to storage systems with siled intelligence or to cloud-anchored SaaS portals without any context regarding your infrastructure environment. Siled systems lack the global context and SaaS portals simply monitor dashboards without integrated AI to predict disruptions based on real-time telemetry data. One-off solutions such as these do not provide an overview.

An AIOps framework integrates IT elements and automates operations, providing an AI-powered infrastructure with cloud agility. Let’s match the essential ingredients with the transport analogy to clearly demonstrate how such an end-to-end framework could work for infrastructure.

  1. Learn to the limit. Much like collecting data from in-car sensors, the business system collects telemetry data from the computing stack – storage, servers, virtualization software, and applications. Most IT vendors provide access to performance logs that can be used to troubleshoot issues once they occur; but it is like taking your car to an auto repair shop after an engine failure. If, on the contrary, your system could learn the local workload patterns in your business and detect any abnormal behavior, it could predict possible failures and help you avoid unplanned downtime rather than sending you to the “shop floor”. repair ”after the fact.

  2. Cloud versatility. To say that the cloud has enabled never-before-imagined possible use cases would be an understatement. Just as the cloud has played a central role in products as sophisticated as autonomous vehicles, it plays an equally essential role in bringing an AI-powered experience to enterprise computing. In addition to scalability and elasticity, the cloud offers vendors several unique advantages: a single, non-siled repository for exploratory data analysis, the convenience of training and iteration across different ML models, and the ability to run simulations and perform multivariate analyzes. These are all essential factors for top-notch AI.

  3. Decision making at the limit. In recent years, edge devices have evolved from simple data collectors to more engaged “decision-making” devices. Edge processing in cars automates changes in vehicle operation – acceleration, cornering, braking – based on real-time sensor data. Likewise, based on the global learning built into integrated AI models, enterprise IT systems continuously monitor impending failure events in real time and ensure minimal or no downtime to your applications around the clock and 7 days a week. This intelligence and continuous decision-making enables systems like HPE Alletra, for example, to guarantee 100% uptime.

The business impact of an end-to-end AIOps framework

As with any AI, such an AIOps framework works behind the scenes. But the benefits specifically tailored to your environment are clearly visible.

  • Improved uptime – Downtime is never good. But the only way a storage system can avoid downtime is to see, learn, and apply that knowledge in ways that anticipate future episodes. By bringing relevant telemetry data to the cloud and correlating it within those datasets, an AIOps platform like HPE InfoSight can provide contextual recommendations to your environment. In this way, an AIOps framework truly improves the uptime of business systems.

  • Business agilityNothing replaces agility in the modern business. Every data-driven business needs IT to meet its ever-changing needs. Ever-changing workload models and an endless stream of new applications can lead to performance bottlenecks and slow response times, unless your data center is equipped to handle everything in its power. submitted. By capturing performance-related telemetry data, an AIOps framework ensures peak infrastructure performance at all times, keeping your business running.

  • Autonomous operations – Arguably the hardest part of AI is not just finding the problem, but coming full circle and automating the recommended action. In most AI-based products, this last step is left to humans. Automating action is the most important difference between a truly stand-alone operation and an operation that simply incorporates AI. Automated action can only be allowed if there is confidence in the AI ​​and precision in the action itself. Such confidence and precision is the result of collecting real-world data and simulating actions in several iterations, spread over years. For example, a storage system with billions of TB of data collected over many years can reliably predict infrastructure operations and perform local, real-time actions related to system saturation and performance based on forecast. workload profile. It can also adjust the priorities of background tasks to minimize the impact on actual customer workloads.

Some AIOps frameworks can automatically identify the best system to deploy an application workload, thereby speeding up application deployment. HPE offers intent-based provisioning, which determines where data should be stored across your entire fleet without the need for storage expertise – with real-time context provided by HPE InfoSight to identify margin for resources and application-specific SLAs.

The industry-leading AIOps platform

You wouldn’t compromise on safety in cars. Why should you compromise critical workloads in your data center? End to end The AIOps framework is now a necessity for business computing. HPE delivers industry-leading cloud operational experience with the HPE GreenLake Edge-to-Cloud Platform to unleash customers’ business agility. HPE InfoSight, the industry-leading AIOps platform, powers this experience. With more than a decade spent leading the way in this new era of stand-alone operations, HPE InfoSight has saved more than 1.5 million downtime for our customers. That’s a lot of “auto repair” time saved, so you can really enjoy the ride!


About Sandeep Singh

sand profile
Sandeep is vice president of storage marketing at HPE. He is a 15-year veteran of the storage industry with first-hand experience driving innovation in data storage. Sandeep joined HPE after working at Pure Storage, where he led product marketing from a $ 100 million pre-IPO fulfillment rate to a public company with more than $ 1 billion in revenue. Prior to Pure, Sandeep led product management and strategy for 3PAR from pre-income to over over $ 1 billion in revenue, including a four-year tenure at HP following the acquisition of 3PAR. Sandeep holds a BS in Computer Engineering from UC, San Diego and an MBA from Haas School of Business at UC Berkeley.

Copyright © 2021 IDG Communications, Inc.


Leave A Reply