Digital transformation of business is driving unimaginable infrastructure scale and data growth.
More importantly, the world is increasingly decentralized – even within an organization.
Different departments and groups often launch their own operations which may overlap.
Over time these various micro-operations converge and become reliant on one-another.
Yet organizations can be in the dark when asked to understand how the systems work together and what dependencies there are.
Often, outages bring these challenges to light.
It predicts in days, months or years when the available capacity will be exhausted, enabling workloads to be moved, balanced or retired, or new capacity added.
“Virtual Instruments is really looking to reinvent infrastructure management, all around applications,” said CTO John Gentry in an exclusive interview.
He explained that the company’s customers told them they didn’t know in advance what was reliant on what in their data center. In addition, determining how applications competed for resources was a significant challenge.
“We looked at this from the perspective of the application,” he exclaimed. “We are at scale and complexity which exceeds human’s ability to understand and act.”
The solution learns where apps live on the infrastructure and which speak to one another using what protocols. This goes into the heuristics engine while determinating the datastore mapping and what processing and database tier they use.
Workloads can be classified as tier-one for something like online transaction processing or tier-three for post-application processing.
Once completed, apps from the same tier can be grouped together to ensure a tier-three application does not significantly slow apps of higher tiers.
At this point, John explained their AIOps solution can determine what is normal on a seasonal basis.
“This analytics works with anomaly detection,” he exclaimed.
It can thus trigger an alert at two standard deviations from normal.
Contention modeling is next – it can see what has and is happening in order to see trends and determine potential conflicts or contention issues. You can subsequently use the recommendations to balance and optimize deployments as needed.
Virtual Instruments recently joined the AppDynamics Integration Partner Program to work together to help enterprise customers optimize the performance and availability of their business-critical applications and infrastructure.
“The big and constantly changing marketplace for IT automation has created an environment that’s hard to navigate for IT teams looking to transform their business with AIOps,” said Brian Paul, head of ecosystem strategy and development, AppDynamics. “The IPP is designed to help enterprises understand the vast array of technologies and capabilities available in this complex space and provide a platform for the optimal solution for AIOps. Our partnership with Virtual Instruments provides enterprises with access to a critical combination of infrastructure and APM capabilities that accelerate the AIOps journey and deliver stellar user experiences.”
John says AIOps is misunderstood – many vendors he explained say they have AI and ML but are actually applying time-based correlation, pattern matching and deduplication.
This is not machine learning and AI, he said.
He said these algorithms need time to learn and pointed to the fact his company started in this space in 2014. As a result, they know what algorithms work with what data sets based on experience.
The next big goal for Virtual Instruments is orchestration and automation… Allowing the machines to make intelligent decisions to remove humans from day-to-to management.
He explained this is automation with governance. The machine applies its recommendations and then compares the outcome to its predictions and then learns and refines the algorithm.
For now, it makes recommendations for higher-utilization or better performance.
The bottom line in the world of AIOps is generally similar – regardless of vendor. In each case, the company is looking to bridge cloud, data center and myriad devices – IoT, servers, storage, compute, etc. to better inform IT management of what is happening. To predict upcoming problems – to better utilize resources and to allow the digitally transformed company to run more efficiently. We are excited about the progress Virtual Instruments is making in the market and also can’t wait to see when the technology is ready to allow the AI to work with even less human intervention.