A New York-based startup is bringing to market today its new DevOps Dashboard, which not only gives visibility into the application stack and its subsystems, but will now also enable users to view defined business metrics via the open-source StatsD data export mechanism.
AppFirst, which was founded in April 2009 and which closed a Series A Javelin Venture Partners-led round of funding this summer, has targeted mid-market companies in online gaming and e-commerce, “Where the application is the business,” according to Pamela Roussos, CMO of AppFirst.
The dashboard is the presentation layer in front of AppFirst’s data collection and aggregation platform. It collects data constantly from multiple sources using a patent-pending technology, rather than doing the more traditional interval polling, which can miss issues that occur only intermittently. The collector installs on an organization’s private cloud, or on virtual, physical or hybrid servers, and works with less than 1% overhead, Roussos said. Data sources include log files, Windows Performance Counters, Nagios plug-ins, and StatsD, according to her. After the data is collected, it is sent to the AppFirst back end, where it is aggregated and made ready for correlation in the dashboard.
“People are still finding out about issues from end users,” said Roussos. “Now, businesses want to be proactive” and find and remedy issues before they are seen in the wild.
The StatsD client library, which provides the generic ability to export data out of an application, can be downloaded from GitHub. Developers put a call in their code to the client library that includes the namespace of the data desired. For example, Roussos explained, “If you care about the number of log-ins, you can put a call there in the code, or if you want the number of searches or failed log-ins, from the dashboard, you can create a widget that will return that data.”
Also today, the AppFirst platform will have the ability to auto-detect application stacks, configure the data sources that are relevant to that stack, and create a dashboard that serves up only that data that is useful for that stack.