Data Science Helps Troubleshoot the Datacenter
Data science is being enlisted to help troubleshoot scaled-up IT infrastructure.
A data science platform unveiled this week by startup BigPanda seeks to use analytics to automate the increasingly complicated task of IT incident management. BigPanda said its platform analyzes the steady stream of daily IT alerts and essentially triages them into high-level alerts.
The startup based in Mt. View, Calif., also announced a $7 million Series A funding round that will be used to accelerate data science product development.
After parsing incidents and generating alerts, the platform is designed to automate the manual processes involved in detecting and fixing IT outages. The goal is to help administrators resolve IT issues faster to minimize downtime.
The startup said its data science approach is a response to dramatic chances in datacenters over the last decade as IT and DevOps teams struggle to manage infrastructure failures. The combination of cloud services and virtualization in datacenters has resulted in the scaling and complexity of operations. The average number of daily alerts in a modern datacenter now numbers in the thousands, BigPanda estimates.
The startup said its data science approach automates the most time-consuming parts of the IT alert response process. Its software-as-a-service platform is said to aggregate alerts from various IT monitoring systems. It then applies data algorithms “to automate the heavy lifting out of incident management.”
The data algorithm consolidates “noisy” alerts so IT administrators can quickly trace the biggest issues such as code deployments and other changes that often contribute to IT incidents. The goal of the data science approach to IT management is to help administrators pinpoint and resolve IT incidents that will only grow in number as datacenters scale and become more complex.
Another IT issue in the modern datacenter, BigPanda stressed, is the fragmentation of monitoring. Multiple monitoring tools from new companies like Splunk and New Relic have replaced tools from larger vendors like Hewlett Packard and IBM. BigPanda estimates that datacenter operators on average now use as many as five different monitoring tools, “none of which speak the same language.”
Hence, the startup argues, the need for data science approaches to correlate alerts while integrating fragmented monitoring tools.
“Only through leveraging data science can IT teams tackle the scale of machines, events and dependencies that must be understood and managed,” Assaf Resnick, BigPanda’s co-founder and CEO, asserted in a statement.
According to one user cited by BigPanda, the time needed to resolve issues was reduced by 80 percent using its data algorithm.
BigPanda was founded in 2012. It has so far raised $8.5 million in venture capital. Its latest funding round was led by Mayfield Fund, with participation by Sequoia Capital. Along with Splunk, the startup said advisers include executives from DropBox, Facebook, Netflix and VMware.