Skip navigation

Finding Structure in an Unstructured Data World

Today’s mission critical apps are dependent on not only new types of unstructured data such as videos and texts, but also new heuristic statistical models and analytic frameworks.

Irshad Raihan is Director of Product Marketing for Red Hat Storage.

The term “mission critical app” used to be easy to define. In the past, most developers or storage administrators may have considered inventory or customer relationship management apps as mission critical. Think: PeopleSoft, or Siebel running on Oracle databases. These applications handled structured data in rows or columns. Everything was nice and orderly, with clearly defined data types that could be easily processed via a relational database and squeezed through a rigorous transformation process for back-end analysis by business intelligence systems.

But we no longer live in a highly structured world, at least as far as data is concerned. Today’s mission critical apps are dependent on not only new types of unstructured data such as videos and texts, but also new heuristic statistical models and analytic frameworks.

These are the new mission critical apps, and we’re seeing them put to use across different industries. Insurance companies are using video and imaging to run claim processing alongside drivers’ histories and calculate premiums using actuarial data. Financial services firms are using modern AI algorithms and predictive analytics along with traditional transactional applications. Manufacturers are analyzing pictures and videos from assembly line robots and Internet of Things (IoT) devices for quality control. And public service agencies using applications to process and analyze data from CCTV footage, texts, and social media in the fight against crime and terrorism.

Unstructured data can be messy, however. This is where object storage comes into play.

Object storage offers some significant benefits for enterprises managing a lot of unstructured data. Data to be more easily found over a distributed system. Stored metadata can go so far as to tell which camera shot a particular video or took a specific photo, and even what actors were used. Object stores are also extremely scalable, and can grow along with the data. In short, highly scalable object stores help bring structure to the unstructured.

The challenge is that not every piece of data in an organization will be unstructured. Most organizations will likely have a combination of traditional structured and unstructured data resources. Facilitating faster and more agile application development will require administrators and developers to be able to easily submit queries across these resources. They need distributed access to a shared storage repository--essentially a shared data context--that allows developers to unlock the data they need, whenever they need it.

Object storage is essential for this type of environment. Object storage allows organizations to decouple compute from storage. Since object stores are built on a flat structure, they offer an ideal foundation for moving from distributing data to distributing access to data. This provides greater flexibility, scale, and durability, as well as the potential for significant cost savings and increased utilization.

Object storage also creates an atmosphere that is critical to managing multi-tenant workload isolation, enabling agility, and preventing data duplication. By combining highly scalable object stores with elastic compute platforms, enterprises can create a distributed data environment that lets analytics teams set up customized clusters as they wish, helping them meet their SLAs without having to recreate or move large data sets. Data platform teams can find consistency among multiple analytics cluster silos. Meanwhile, data engineers can dynamically provision those clusters with the right resources, versions, and data.

While businesses may like structure, it is unstructured data that is driving today’s businesses. Organizations need a better way to store, manage, and distribute this data. Providing multi-tenant workload isolation with a shared data context is the key. This can be achieved with a combination of  highly scalable object storage and flexible compute platforms enabling IT administrators, cloud operators, and application developers, to curate the data they need for today’s mission critical apps.

Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Informa.

Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.