With the rise of AI, storage management has evolved from storage resource management software to AIOps tools that automate large parts of the process.
AIOps enables monitoring, diagnostics, predictive analytics, and prescriptive capabilities for storage infrastructure and applications. Essentially, AIOps tells organizations what's happening with their storage, why it's happening, what could happen, and what to do about it.
AIOps increases efficiency by eliminating much of the manual work of storage management, freeing up IT staff to focus on other tasks. Various vendors offer AI for storage and storage management, each with their own strengths and weaknesses.
How AI storage management works
AIOps for Storage uses machine learning to collect and analyze telemetry data, which is then transformed into predictive analytics.
AIOps encompasses automation, performance management, and service management, automating many of the decisions involved in scaling and securing storage systems. Additionally, AI can help with tasks such as storage planning, storage lifecycle management, root cause analysis, and storage optimization.
A key component of AI storage management tools is telemetry data: information collected through sensors from storage, server, and network systems.
AI and machine learning analyze information collected from these systems across hardware devices, OS, applications, and hypervisors, allowing detection of anomalous activity such as misconfigured devices, unexpected capacity growth, and unusual throughput demands, which can be used for resource planning and storage performance optimization.
AI storage management systems are often SaaS applications that run analytics in the public cloud, allowing them to compare that data with information collected from various systems to improve their predictive capabilities, although several vendors offer products that are specifically designed to run on-premises.
What differentiates AI-based storage management tools is their functionality. Many of these tools use AI to assist with troubleshooting, root cause analysis, and storage optimization. Some tools offer additional capabilities, such as assisting with power consumption, application deployment, and even hardware lifecycle management.
While most AI-enabled storage tools are proprietary in nature, some storage vendors are leveraging APIs as a way to collect data from third-party sources.
Before investing in an AI storage tool, consider how the system delivers alerts. Most tools display alerts on a dashboard, but some use other mechanisms, such as text or email messages. Whatever the delivery method, the tool should filter out unnecessary alerts so your organization can focus on what's important.
AI Storage Management Options
Several vendors offer AI storage management software that is compatible with a variety of systems and applications. Vendors are also using AI storage management to offer storage as a service.
Vendors and the systems and applications they cover include:
Dell Apex AIOps Infrastructure Observability. All Dell storage, PowerEdge servers, VxRail, PowerFlex, and VxBlock converged and hyperconverged infrastructure (HCI), PowerProtect Data Domain and PowerProtect Data Manager data protection, PowerSwitch and Connectrix networking.
HPE Info Site. Alletra, Primera, Nimble storage, SimpliVity HCI, ProLiant and Apollo servers, Synergy composable infrastructure.
IBM Storage Insights. All IBM block storage, switches, fabrics and VMware ESXi hosts – the paid subscription version of IBM Storage Insights Pro also covers IBM and third-party block and object storage.
Infinidat InfiniVerse. InfiniBox, InfiniBox SSA.
NetApp Active IQ. OnTap, Element, StorageGrid, SANtricity.
Pure Storage Pure1. FlashArray, FlashBlade, and Portworx storage.
The Benefits of AI Storage Management
AI-driven storage management eliminates much of the complexity and manual tasks of traditional storage resource management. Benefits include automated provisioning, intelligent data tiering, and workload optimization. As a result, dedicated storage personnel can spend less time monitoring and managing systems. AIOps is a great help for MSPs, as it allows them to remotely manage storage for many of their customers.
Management systems can help users prevent unexpected problems by predicting future events based on current usage patterns. They can also advise users to add storage capacity, compute resources, and complete other upgrades before performance becomes an issue. Users can configure applications to automatically take actions to prevent device failures or poor performance, but IT departments may prefer to receive recommendations and make changes themselves.
AI Storage Management Challenges
One of the problems with AI storage management systems is that they are often proprietary and usually only work with one vendor's products. For example, if an organization uses a SAN from one vendor, an AI storage management system from another vendor may not be compatible with it.
Another issue is that data collection and analysis creates more data, which in turn creates more storage requirements for storage management. Over time, organizations must decide which data to safely discard.
Not all organizations can afford to allow third parties to connect to their data centers. So-called dark sites cannot use SaaS-based analytics that collect and store data in the public cloud or at the vendor's site. Vendors can use analytics software on local servers so that telemetry data does not travel, but these users lose some of the benefits of analytics because their data is not compared to that of their peers.
AI-based storage management will evolve as a direct result of AI technology maturing.
All users, dark site or not, should ask their vendors how much information they are collecting outside of storage and how they ensure that data remains anonymous and protected.
AI in storage management is still a relatively new technology, so algorithms will improve over time as more information is collected.
The Future of AI in Storage Management
AI and machine learning technologies have advanced exponentially over the past few years, so AI-based storage management will evolve as a direct result of AI technologies maturing.
Storage vendors are more likely to use AI as a security tool: For example, companies can train AI monitors to recognize the signs of a ransomware attack and prevent potential infections.
AI-based storage management tools are more likely to support intelligent backup and recovery capabilities: For example, AI can automatically identify an organization's most important data, ensure that it is backed up, and prioritize this high-value data if a restore is required.
AI in storage systems can also provide self-healing capabilities, allowing the AI to detect disk failures, corrupted sectors, or similar issues and take corrective action to prevent data loss or system outages.
Brien Posey is a 22-time Microsoft MVP and commercial astronaut candidate with over 30 years of experience in the IT industry, including serving as a Principal Network Engineer for the US Department of Defense and a Network Administrator for one of the largest insurance companies in the US.
Dave Raffo worked at TechTarget from 2007-2021 as executive news director and editorial director for storage.
