Application Management Gets Unraveled
It’s all about enterprise applications, we are told, with big data apps among the most critical. Hence, a growing focus on managing application performance has fueled new monitoring approaches such as operational data science.
The latest “full-stack” approach comes from Unravel Data, which this week launched an application performance monitor (APM) platform that supports both Cloudera Impala and Apache Kafka. Impala is the SQL query engine that runs on Apache Hadoop. Meanwhile, Kafka has emerged as a key stream-processing platform for handling real-time data feeds.
Unravel Data, a startup based in Menlo Park, Calif., said Tuesday (June 6) its new platform addresses the growing complexity and time required to diagnose and resolve growing application performance problems. The startup’s “application-first approach” seeks to provide a unified view across many components in a technology stack, including those underlying applications such as machine learning and analytics.
“People don’t usually understand what’s happening in the stack, and this impedes its performance,” Kunal Agarwal, the co-founder and CEO of Unravel Data, told Datanami last fall.
The platform is designed to tune and speed up Impala queries, for example, by tracking down the root causes of slowdowns and autonomously implementing steps to boost performance. Memory allocation issues and other resource contention issues are among the issues addressed.
The company also said its platform can be used to identify SQL queries from engines such as Apache Spark that are best suited for running Impala along with those that are not.
Meanwhile, growing adoption of Kafka across enterprise infrastructure has also increased requirements for monitoring performance at scale, the company asserts. The stream processing platform is increasingly being used for big data applications such as ingesting data into clusters and delivering live data feeds for used for real-time processing.
Along with performance issues at the application level as well as within multi-tenant infrastructure, the startup notes that performance problems can arise within Kafka itself.
The Unravel Data platform is intended to manage applications based on Kafka by correlating application performance indicators with underlying infrastructure along with Kafka performance metrics. It also tunes applications that are unable to keep pace with Kafka input data rates.
Along with the proliferation of big data applications extending form Hadoop and Spark to Impala and Kafka, Agarwal noted “the growing number of interactions between apps on the stack, which are multiplying almost geometrically.”
He added: “Something will break—in fact, it’s bound to break. Big data practitioners come to the table with that understanding, and yet they have no tools to cope with that complexity. And the challenges continue to multiply with the exponential growth of data in most organizations.”
Unravel Data said the 4.0 version of its platform for tracking application performance running on Impala and Kafka is available now.