Create reports and notify teams when performance issues arise.
Get started
Service Level Objectives (SLOs) are the established method to track system performance in SRE and DevOps organizations. Each SLO defines a target value for an SLI (Service Level Indicator, a defined metric). For example, a performance SLO for a web application might state that 99% of requests must be answered in less than 500 ms each hour. Each hour in which the objective is not met counts against a defined "error budget". When the error budget has been depleted, actions need to be taken to improve performance and get back into SLO compliance.
SLOs can be difficult to manage when many metrics across different sources of monitoring data are involved.
Use time-series data from your favorite tools to feed SLO watchers and reports.
Get StartedFor each tool you want to access through the Metralyze API, you need an account or API token in that tool. For example, for Instana you will need to create an API token.
All configuration in Metralyze is available via API. That means that watchers, SLO definitions and connections be managed from within CI/CD pipelines.
Metralyze focuses only on time-series data, which is characterized by having a set of time-stamped metrics together with a number of dimensions that give context for those metrics. Many monitoring and analytics tools offer the same kind of data. What we do is to offer a common query API and data format that works across all tools that we integrate with.
Yes, Metralyze is already being used by companies to monitor time-series data and report on SLOs. However, the product is currently in "private beta", meaning that we are only accepting a small number of additional users as we scale up the infrastructure. Contact us to discuss your use cases.
Contact us for options on how to set up connections to time-series APIs that we haven't connected yet.
"There is nothing permanent except change" - in an ever-changing world of digital products and services, time-series data is essential for monitoring changes in behavior of humans and machines. "Time-series data" is what we call series of time-stamped records that reflect how certain measurements (metrics) change over time, and where those measurements were taken (dimensions). IT monitoring tools capture massive amounts of time-series data. Web analytics tools capture time-series data. IoT systems capture time-series data.
Time-series data is everywhere, but it's under-used: There is often too much of it for humans to process, and the data is most often siloed inside of disparate tools. That's too bad, because all time-series data has one thing in common: time! Using timestamps, we can correlate changes in one area (for example, customer conversion rate) with changes that happened to take place at the same time (for example, an increase in page load times or error rates).
With Metralyze, we want to change that: by connecting systems that collect time-series data and making it easy to access that data, we open the door to advanced analytics across different functions:
