Big Data

For a better cloud strategy, look to data

Share this post:

Cloud strategy dataEnterprise cloud strategy is undergoing revolutionary changes.

While initial deployments focused largely on back-office support functions and migrating workloads to off-site infrastructure, the modern cloud has become a go-to solution for top-tier workloads and forward-looking initiatives, such as mobile data and the Internet of Things (IoT).

While most people have traditionally viewed the cloud as a warehouse for generic computing, storage and networking resources, that model is losing ground. Increasingly, enterprise cloud environments reflect a highly optimized and customized infrastructure.

Attaining this state of digital nirvana can seem imposing, since complex cloud environments with both public and private components require more oversight to manage. Luckily, the key to optimization is already at your fingertips: It’s your data.

Better infrastructure through data

With the right data collection and analysis, companies can very easily determine what’s needed to maintain optimal performance and begin shifting workloads to the appropriate resources, even for highly dynamic use cases.

According to Ken Christiance, distinguished engineer on the IBM Technology, Innovation and Automation team, one of the key elements of an optimized cloud is proper workload balance. This can be achieved through a capacity management analytics (CpMA) solution, such as the Watson-powered Densify platform. Using deep visibility and detailed reporting tools, Densify can identify problems such as over- and under-allocated virtual machines, inefficient workload placement, imbalances between hosted resources and application needs, and inadequate cluster capacity. This helps infrastructure managers to stay on top of working conditions within the cloud, or, even better, feed this performance data into an automation engine that continuously fine tunes the environment for optimal performance.

A recent post by Tony Efremenko, executive architect at the IBM Cloud Garage, highlights how this works in practice on another Watson-powered platform called Studio. His level of optimization extends beyond merely tweaking infrastructure to meet current demands. It helps build predictive models that can determine what will be needed in the future, too. For example, microservices tend to be highly dynamic and can be employed in myriad ways. However, by putting Watson Studio to work, supported by real-time data collection, Efremenko can establish the right cost structure for multiple services regardless of whether they were developed in-house or on third-party software-as-a-service (SaaS) platforms.

AI in the mix

The increasing complexity of modern cloud environments all but demands that the enterprise uses artificial intelligence (AI) platforms such as Watson in the management stack sooner rather than later. James Kobielus of Silicon Angle noted recently that a successful multicloud strategy depends on the ability to automate multiple tasks, including log analysis, anomaly detection, root-cause diagnostics and closed-loop issue remediation. If these functions remain manual, performance will surely suffer and costs will become prohibitive. Going forward, look for AI to work its way into both infrastructure and application-layer management.

One of the lesser-known challenges of AI, however, is the tendency for operator bias to creep into learning algorithms. When applied to cloud management, this causes decisions regarding resource allocation and other functions to veer away from true optimization and more toward administrative preferences. IBM is helping the enterprise combat this issue with a new Trust and Transparency Service that detects bias and sheds light on the decision-making process. The system works with all leading machine-learning frameworks and features a number of customization tools to tailor performance to specific enterprise clouds.

AI is only the latest advanced technology to infiltrate distributed virtual infrastructure, but by no means will it be the last. To help navigate these uncharted technological waters, enterprises will want to partner with a proven technology leader that is not only at the forefront of cloud infrastructure, but has a vision for the future as well.

Properly making use of data that already exists within the cloud environment is key to ensuring enterprise cloud spending is producing the highest return. As data users become more accustomed to getting what they want when they want it, the only way to remain relevant will be to put in place a highly optimized, customized cloud ecosystem.

Learn more about how IBM can help you create a hybrid cloud environment that’s purpose-fit for your enterprise. Read how IBM can help create the right infrastructure for your big data strategy and improve performance. Register for the full report on infrastructure.

More Big Data stories

Accessible bookcast streaming service becomes securely available on IBM Cloud

Accessible entertainment isn’t often easily accessible. The mainstream entertainment industry doesn’t cater to people who can’t see or hear well; people with PTSD, autism or epilepsy; or those who are learning English as a second language. These population segments can become isolated, and then marginalized when they can’t enjoy the latest feature film – even […]

Continue reading

QAD turns cost centers into profit centers with IBM Cloud

Cloud has been an important part of our strategy at QAD for well over a decade. In fact, among the established global manufacturing enterprise resource planning (ERP) and supply chain software providers, QAD was one of the first to offer cloud-based solutions, starting with QAD Supplier Portal in 2003 and then ERP in 2007. Despite […]

Continue reading

How a hybrid workforce can save up to 20 hours a month

How productive would your company employees be if they could save two hours a day on regular tasks? With the growth and evolution of today’s digital economy, companies face the challenge of managing increasingly complex business processes that involve massive amounts of data. This has also led to repetitive work, like requiring employees to manually […]

Continue reading