SignalFx announced the general availability of its latest release, featuring new alerting capabilities that empower cloud operations teams to better monitor and manage cloud infrastructure, containers, and applications.

Monitoring cloud applications requires ingesting and analyzing data from hundreds to thousands of web services, many of which employ scale-out, elastic architectures with highly variable workloads. Determining the best alert conditions in those environments, and the impact new alert conditions will have on the operations team, is a complex process. SignalFx removes the complexity and maximizes the productivity of the cloud operations team with powerful new tools that expedite the creation, deployment and tuning of alerts using machine learning algorithms that dynamically adapt to changing environmental conditions. Now cloud operations professionals can preview the alert conditions on historical data, leverage prebuilt advanced alert conditions for cloud applications such as anomaly and outlier detection, and quickly build custom alerts through the SignalFx API and alert condition library.

“Running a cloud-based platform means we have to account for all of the possible risks to our business operations while monitoring thousands of instances and applications.” Said Jay Ferrin, Vice President, Quality & Infrastructure Services at Acquia. “The new alerting capabilities from SignalFx will ensure our engineers are able to configure the alerts they need to quickly identify the source of performance and stability issues so we can minimize the impact they have on our customers.”

Key features of today’s announcement include:

■ Alert Preview: SignalFx now provides cloud operations professionals the ability to test and preview the results of alert conditions on historical data. Users will be able to see the frequency and efficacy of alerts before applying them to live, real-time data streams, removing the cost and confusion of unnecessarily firing alerts.

■ Built-in Alert Conditions: SignalFx now includes pre-built alert conditions specifically designed for cloud operations, saving customers time when setting up their monitoring and delivering alerts that reflect the reality of operating cloud applications. These conditions include:
- Outlier detection which, for example, detects when a load balancer misconfiguration results in a disproportionately high traffic pattern to a subset of a cluster.
- Sudden change alerting where sudden changes are detected based on comparison with recent history, such as when a configuration change or code push results in a sudden acceleration in an API latency metric.
- Historical anomaly detection where a signal differs by a specified amount from the established norm for a given time period, for example, or where there is a sudden spike in use of a specific microservice tied to a revenue generating process that is far outside of the average load for that service on a Monday morning between 10 and 11 a.m.

■ Alert Functions Library: The sophistication and complexity of cloud applications requires operations teams to develop custom alerts that reflect the needs of their environments. SignalFx now delivers deeper operational intelligence in a much easier way by providing direct access to the alert condition library via the SignalFx API. The SignalFx API extends all the capabilities within the SignalFx UI and enables users to automate new operational use cases.

These new alerting features are built on SignalFx’s industry leading streaming analytics technology for time-series metrics. Unlike most other monitoring solutions that evaluate alert conditions against data once fully collected and stored in a database, SignalFx applies analytics and evaluates alert conditions against data as it arrives in real time. This novel approach enables SignalFx to build sophisticated alert conditions that detect and fire within seconds of a condition being met across populations of thousands of instances. Whereas most monitoring solutions take minutes to alert on simple conditions, SignalFx can alert within seconds of a more meaningful pattern emerging.

“Any organization moving to microservices or leveraging container based architectures is faced with new challenges around managing alert noise and minimizing triage time when something is going wrong with their application,” said Karthik Rau, CEO and co-founder of SignalFx. “We are excited about how this new release democratizes the benefits of machine learning to everyday users of monitoring systems, substantially reduces alert noise, and enables operators to proactively identify emerging issues before they impact end users.”