Shuo - stock.adobe.com
AIOps and storage management: What it is and who provides it
We look at AI-powered tools to manage storage and data protection that can help to detect infrastructure issues and help to optimise performance and even cloud costs
Artificial intelligence (AI) is an obvious candidate to take some of the load off IT administrators. By using AI to automate repetitive tasks, IT teams can manage more infrastructure more efficiently and with less risk of human error.
AI systems can react more quickly than humans to infrastructure faults and to manage capacity and performance, as well as alert admins about data protection issues and even help to optimise costs in cloud deployments.
Such AIOps tools now play a growing role in storage management. Most storage suppliers have embraced AI-based tools to make use of the large volumes of management data produced by storage systems and the quick wins that come with better storage management.
In this article, we take a look at the key suppliers’ offerings.
Storage suppliers’ AIOps tools
Enterprise hardware suppliers have increasingly added AIOps features to their IT management tools. They exploit telemetry connections to hardware and machine learning (or AI) to process the vast volumes of data, alert administrators, or even attempt a fix.
But supplier approaches vary, with some that integrate storage management into general purpose toolsets and others that focus on specific use cases such as the datacentre.
There is also an emerging market for third-party and open source AIOps tools.
Dell
Dell’s APEX AIOps provides SaaS-based observability and management across IT systems. APEX offers infrastructure and application observability, as well as IT incident management. The tool works across Dell’s core on-premise infrastructure, edge and multicloud environments, as well as its hyper-converged infrastructure.
APEX provides system health analysis, capacity and performance anomaly detection, performance forecasting, and intelligence on security risks. APIs allow APEX to work with other tools. It works with all Dell storage.
Hitachi Vantara
Hitachi Vantara’s Hitachi Ops Center includes AI-driven system configuration tools, performance analytics, workflow automation and observability, as well as data protection.
Hitachi Ops Center focuses on datacentre management, using AI and machine learning for forecasting, performance optimisation and to predict faults. Ops Center can manage and optimise virtual machines and collect data from multiple storage systems. Hitachi Ops Center Administrator sets out to reduce storage administration across Hitachi’s Virtual Storage Platform.
HPE
HPE InfoSight AI supports the supplier’s Alletra, 3Par, Primera and Nimble storage products. InfoSight is free for users that have a support contract in place. It uses APIs to automate hardware management, including that from HPE partners. This includes downtime prediction and prevention, capacity forecasting and signature detection.
InfoSight works with HPE’s OneView IT management tool for software-defined infrastructure.
Huawei
Huawei’s Data Management Engine (DME) is an operations and maintenance platform. This includes performance prediction and fault identification. DME checks for correct load balancing and can carry out root cause analysis of failures. Huawei AIOps also includes options for cloud management and for virtualised environments.
IBM
IBM provides its Storage Insights tool for block storage management to subscribers to Storage Insights Pro. It also supports some storage technologies from Dell EMC, Hitachi Vantara, NetApp and Pure Storage.
IBM Turbonomic brings a range of datapoints together for cloud cost optimisation – this in turn works with Instana Observability. Instana also provides automated resource allocation and optimisation, as well as AI-powered incident management.
Pure Storage
Pure1 AIOps is a SaaS storage management platform that includes storage optimisation, Purity (operating system) upgrades, and subscription management for on-demand storage.
Pure1 includes storage planning and internal chargeback calculations, workload placement and lifecycle management. Pure1 comes with all the vendor’s subscriptions and support contracts.
Pure also plans to launch an AI copilot for storage, which will include features such as threat detection, as well as storage performance management.
NetApp
NetApp Keystone Storage as a Service (STaaS) is based around AIOps. This, the supplier says, allows users to build “cloud-like” functionality with on-premise storage hardware.
This uses NetApp BlueXP digital adviser, which in turn runs on NetApp’s Active IQ Digital Advisor. The tools provide usage and capacity prediction, and security and anomaly detection. These come from NetApp’s Data Infrastructure Insights, formerly Cloud Insights. Keystone STaaS also provides subscription alerts for cost optimisation for on-demand storage.
Outside of Keystone STaaS, ActiveIQ automates storage health and cost optimisation as well as security and resilience to other NetApp products. The Digital Adviser covers infrastructure optimisation, while an operational resiliency dashboard works with on-premise ONTAP clusters to mitigate any vulnerabilities.
As well as Keystone and ONTAP, NetApp Active IQ also works with E-Series and StorageGRID storage.
Third party, open source and cloud AIOps
Supplier AIOps tools promise tight integration with their storage hardware and often other parts of their stack. But, cross-vendor compatibility is more limited or often not supported at all.
A number of third-party AIOps tools aim to fill the gap for multi-vendor environments. One example is Digitate’s ignio, a SaaS-based tool that provides automation, observability and fault finding.
The open source community has also developed AIOps tools, although many of these are components rather than complete AIOps toolsets. These include AIOpsTools, Grafana and Prometheus.
The cloud hyperscalers also offer some AIOps capabilities. Although it does not specifically target storage, AWS’s Amazon Lookout for Metrics, for example, provides performance monitoring and anomaly detection for AWS and third-party cloud apps.
Read more about storage management
- Storage explained: Consumption models of storage procurement: We look at consumption models of storage purchasing and how cloud operating models have made them mainstream and supplanted the traditional three-year lift-and-shift datacentre refresh.
- Storage performance metrics: Five key areas to look at: We look at the essentials in storage performance metrics: capacity, throughput and read/write capability, IOPS and latency, and hardware longevity measured by failure rates.