Big storage for big data

1. Summary

While the lure of the free technology that open-source software projects offer for large-scale object storage can be attractive, IT organizations must recognize the potential risks that they will be taking and, even more, the significant cost commitments.

With large and distributed data sets, storage for big data is a huge challenge. Data can be collected, stored, and distributed from in-house, on-premise, and in-cloud sources. Enterprises that want to take advantage of big data must set the stage with a framework that enables large-scale object storage systems to be deployed. CIOs and data center architects must work with software developers and independent software vendors (ISVs) to deploy an underlying infrastructure that is flexible, scalable, and predictable.

As an aid to those decision-makers, in this report we’ll review the impact of open-source software projects on infrastructure, cost, and support/maintenance and then recommend an approach and strategy that will work best for most enterprises.

Key findings include:

• While open-source storage software offers many advantages, potential customers should balance the criticality of their enterprise data against the emergent nature of open-source solutions.
• A direct link exists between building enterprise value and the competitive advantage big data offers, but there is not necessarily a benefit from “tinkering” with a mature computing infrastructure.
• A major drawback of open-source solutions is that while the software is free, the hardware is not, and neither are support and maintenance, which can total 70 percent of an IT budget. Capex may fall marginally, but opex could rise dramatically.
• Ultimately, open-source software, particularly as it relates to object storage systems, is best used when the open-source project has a specific feature that addresses a glaring issue lessening the enterprise’s ability to generate value or competitive advantage.

Thumbnail image courtesy of: iStock/Thinkstock.