As demand for digital services increases and distributed systems become more complex, organizations must collect and process a growing amount of observability data (logs, metrics, and traces). Site reliability engineers (SREs), developers, and security engineers use observability data to learn how their applications and environments are performing so they can successfully respond to issues and mitigate risk.
With use cases expanding across many business units, it's important for organizations to know how users in various roles use observability data. A new report from The Harris Poll and Mezmo explores this concept. Based on a survey of 300 SREs, developers, and security engineers in the US, the study digs into key pain points and how companies might use observability pipelines to help make decisions faster.
Observability Data Is a Part of Daily Usage
More than half of SREs, developers, and security engineers use observability data daily, with another third of people in each role using it two to three times per week. Typical machine data interaction looks different for each role. SREs focus on troubleshooting, analytics, and monitoring uptime; developers on troubleshooting and debugging; and security engineers on cybersecurity, firewall integrity, and threat detection.
The Amount of Data Is Escalating
Data volume is increasing considerably and becoming difficult to control as data is spread across many systems and apps. While respondents in all three roles use a median of four data sources to get their jobs done, SREs and developers often use three separate products to access that data, and security engineers use two. And over the last 12 months, developers and security engineers have seen a median of two new data sources being added, and SREs have seen three.
Adding new data sources and controlling the flow of data has become an overly complex process involving many different tools that don't integrate well and provide delayed insights. Organizations must harness all this data to make real-time business decisions because a slight delay can cause issues.
Difficult to Control Skyrocketing Costs
In addition to data volume, the three groups listed cost control as a top challenge. Specifically, 92% of SREs, 99% of developers, and 97% of security engineers say it's hard to manage the costs of collecting and storing data. High volume of data creates budget pressures across the organization as budgets are not increasing proportionally to the cost. Organizations must look for ways to extract more value from their telemetry data by making data available to wider teams for additional use cases. This requires free flow of usable telemetry data to any platform of choice.
Making Data Actionable with Observability Pipelines
Most professionals in all three roles agree that newly adopted technology, like observability pipelines, must integrate with existing data management platforms. When looking at observability pipelines to help better control and take action on data, all three roles report that supporting cloud data sources is essential. SREs and developers are also interested in making sure that cloud application data sources are supported, while SREs and security engineers need to be sure that there is firewall data source support. However, teams are not just looking for collecting data but need various transformations to add additional context to the data. They are looking for capabilities such as log transformations, sampling, enrichment, and augmentation to make data more meaningful and actionable.
As the report reveals, the importance of observability data is growing, but organizations are challenged with making this data actionable. Observability data pipelines are an emerging technology organizations can use to collect, transform, and route all this data to various teams for greater actionability. Once organizations can understand how different groups use this data, they'll be able to extract greater value for the business.