sdecoret - stock.adobe.com

Can AI save the storage admin?

AI initiatives have created a surge of data, making traditional storage management insufficient. How much can AI help solve the issue it is creating?

Given the current surge in enterprise data spurred by widespread investment in private AI initiatives, we must acknowledge that traditional storage management is broken. Data has long been vital to business operations. But now, in the AI era, data must deliver valuable insights to unlock greater efficiency in all areas of the business while fostering customer engagement and generating new revenue streams.

Storage administrators face demands for greater scale and control over ecosystems that span multiple locations and cloud providers. Data locality, sovereignty, security and performance are all becoming higher priorities. Success for data storage administrators is no longer just about ensuring cost-effective capacity and the protection of data; it is also about being superior data stewards, ensuring the optimal performance, optimization and control of data while providing the right teams with the right data when they need it. Despite these increased demands, storage administrators are not getting the help they need in the traditional sense.    

According to recent research from Omdia, a division of Informa TechTarget:

  • 69% of organizations agree that the complexity of their organization's IT infrastructure environment slows IT operations and digital initiatives.
  • 68% agree that data storage is a major pain point for their organization.
  • 85% of storage administrators agree that they have taken on new responsibilities (or are under pressure to) to support their organization's digital transformation goals or initiatives.

While the rise in private AI investment is a catalyst for this need to rethink storage management, these complexity concerns are not localized solely to AI workloads or enterprise organizations pursuing private AI. The complexities of managing existing workload infrastructure, ensuring consistently optimized performance, ensuring availability and securing the environment steal cycles from administrators and create an opportunity cost to practically every business.

Traditional simplicity is not enough. For decades, storage vendors have promoted their traditional simplicity features. Every system is easy to use, and yet complexity increases every year. The challenge is that, at scale, managing a vast multitude of disparate "easy" systems creates compounding levels of complexity.

Is AI to the rescue?

Recently, nearly every major enterprise storage vendor, including Dell Technologies, HPE, IBM, Hitachi Vantara, NetApp and Pure Storage, has begun offering integrated AI capabilities within their infrastructure management and/or observability tools. The integration offers multiple opportunities to offload menial or mundane tasks from administrators. AI can offer provisioning and optimization recommendations; automate actions such as data provisioning, migration, or the application of data services; identify compliance lapses; and help diagnose the root cause of performance concerns or other issues.

For storage buyers, evaluating the benefits of these integrated artificial intelligence capabilities must be a priority moving forward with new investments, given the potential for time savings.

Some of these capabilities deliver real and valuable time savings, while other features provide another path to possibly the same outcome. For example, the ability to instantly diagnose and recommend fixes for performance concerns likely provides significant time savings, while asking the AI prompt to create a 10TB volume versus using existing scripts might not move the needle as much.

For storage buyers, evaluating the benefits of these integrated artificial intelligence capabilities must be a priority moving forward with new investments, given the potential for time savings.

Last week, for example, IBM announced FlashSystem.AI, comprising new integrated AI capabilities for its FlashSystem enterprise storage portfolio. While multiple vendors offer similar AI capabilities for storage management, IBM's recent FlashSystem.AI demonstration acts as an example of a tool possessing features that can improve the time-saving potential of AI.

  1. View Rationale -- When the FlashSystem.AI makes a recommendation, such as where to provision a new volume or how, the administrator is provided a "view rationale" option. When the administrator selects this option, IBM's FlashSystem.AI automatically provides the charts and metrics used for the recommendation. The immediate presentation of the background rationale should improve the confidence of storage administrators, accelerating their adoption of the recommendations. In addition, this data also simplifies cross-team collaboration by automatically supporting the storage team with the justification behind the recommendations, if or when different tools in use by different teams have conflicting findings.
  2. AI-powered non-disruptive data migration -- By using FlashSystem's integrated storage virtualization capabilities, FlashSystem.AI can recommend and facilitate non-disruptive migrations between FlashSystem systems, while also managing external heterogeneous storage. Continued performance optimization often requires data movement as demands change over time. While storage systems have provided agility features for some time, the collection and analysis of data to justify data movement often add another hurdle, stealing admin cycles and slowing optimization efforts. The integration of AI should make existing agility capabilities, such as those from IBM FlashSystem, more accessible and ultimately more valuable.

With the bulk of the major storage players integrating AI functionality, it is imperative that IT decision makers include these capabilities in the storage evaluation process. Most importantly, focus on AI capabilities that can offload the most time-consuming activities, the breadth of operations available to be optimized by AI and the ability of the integrated AI to quickly instill confidence within the organization.

Scott Sinclair is practice director with Omdia, covering the storage industry.

Omdia is a division of Informa TechTarget. Its analysts have business relationships with technology vendors.

Dig Deeper on Storage architecture and strategy