VMware's had its eye on artificial intelligence for some time now and for good reason. Artificial intelligence...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
has the ability to predict system needs, automate important but tedious tasks and build simulations to help identify data center issues before they occur. VMware has already started to incorporate artificial intelligence into some of its offerings, including vRealize Log Insight and the recently released AppDefense.
First, a clarification: Most examples of VMware using AI are actually examples of machine learning. Machine learning is a type of AI that processes data with statistical analysis algorithms without the need for explicit programming. Let's say you wrote an if-then program to highlight issues in your system. According to this program, if a log entry contains the word error, then there is a problem with the system. Although this is useful, it would take too long to investigate each and every log entry across all of your systems for instances of the word error to actually do your job. VMware vRealize Log Insight uses machine learning algorithms through its Intelligent Grouping feature to scan incoming data and then group messages together to locate events across systems that point to a specific problem.
You might think that, since it's been around for such a long time, AI would play a larger role in more software. The origins of AI date back to the 1950s, but there haven't been any real breakthroughs in the field until recently. That's because, in the past, there was an insufficient amount of compute for AI to work with and a lack of access to the huge amounts of data from which machines need to learn. These days, we have access to both the compute power and data necessary for AI to mature.
The more data, the merrier
Much like human intelligence, AI -- specifically machine learning -- collects data from the surrounding environment and analyzes that data to determine right from wrong, good from bad and so on. However, unlike a human, who can learn from small chunks of data, machine learning algorithms only get better at analyzing data if it receives massive amounts of data input.
Some machine learning-based products, such as vRealize Log Insight and vRealize Operations, only collect and process data from your own environment. Another VMware feature, Workspace One Intelligence, also uses data collection to create more secure and robust systems. The product of VMware's acquisition of Apteligent in May 2017, Workspace One Intelligence collects data from all devices, applications and users in a Workspace One deployment and evaluates that data with machine learning algorithms to improve security and performance.
Other similar products require significant amounts of data from numerous sources. This is where machine learning runs into trouble. Customers understandably value their privacy, but if machine learning-based products don't have access to customer information about systems usage, errors and statistics, their algorithms will have a difficult time drawing correlations between events and becoming proactive.
If you've installed a VMware product within the past few years, then you're probably familiar with the dialog that invites you to join VMware's Customer Experience Improvement Program (CEIP), shown in the figure below. CEIP is the in-product method VMware uses to collect customer usage data. Many administrators opt out of CEIP because they haven't taken the time to discuss privacy issues within their organization and prefer a safe-rather-than-sorry approach. But, if more administrators took the time to research CEIP and opt in, all VMware users would benefit.
The sky's the data collection limit
Although it continues to use CEIP to collect customer data, VMware recently introduced a new proactive support technology for data collection called Skyline. Skyline is currently only available to customers with Premier Support in North America and is expected to become available to Production Support customers in 2018. At present, there's no information about when Skyline might be available to other geographic regions. In its current version, Skyline analyzes vSphere and NSX environments; VMware will add other products over time.
To use Skyline, the customer must install the Skyline Collector virtual appliance in her environment. The Skyline Collector uses automation to collect and analyze configuration, feature and performance data from the environment and then send the results of that analysis to VMware. VMware then processes this information with machine learning algorithms, and support engineers use the information to proactively engage with customers to eliminate existing issues and prevent other potential issues from occurring.
The thing that sets Skyline apart from other similar data collection technologies is that it aggregates data from all customers into one big data lake that engineers can use to analyze and identify trends and events within the entire VMware landscape.
AI is playing an increasingly larger role in VMware's product portfolio, and as the company collects more data over time, its machine learning algorithms grow increasingly more robust, especially as VMware cloud products make it easier to obtain metadata about the use and operations of cloud services. This growth will only continue as VMware expands into the internet of things with its connected devices and sensors that make widespread data collection possible.
A beginner's guide to machine learning
Vendors turn to AI to overcome budgetary constraints
VMware analytics expand to end-user computing