Alerting is fundamental to monitoring software. If you aren’t proactively alerting on the software you manage, then you’ll spend a lot of time in the war room reactively fighting fires when issues arise. That’s why we created New Relic Alerts, to help teams proactively monitor the software and infrastructure that powers their digital business.
We take service quality very seriously, so we baked New Relic Alerts in beta for some time. After the beta period, we carefully rolled it out to approximately 30,000 accounts, and now we’re ready to throw open the gates and release it into the wild. So we are super excited to announce that New Relic Alerts, a core component of the New Relic Digital Intelligence Platform, is now generally available to all paying New Relic customers!
Alerting as a craft
Good alerting is like a craft: a combination of art and science. New Relic Alerts provides the tools to help with that craft. The science comes from the instrumentation and data you use New Relic to collect across the full technology stack of your business; the art comes from how all the knobs and dials provided by New Relic Alerts help you create meaningful and actionable alerts from that data. New Relic Alerts is designed to give control back to the on-call teams supporting complex software, so that they know if their infrastructure, apps, and ultimately their digital businesses are running smoothly.
To enable that, New Relic Alerts was created in accordance with four key principles:
- Dedicated, centralized UI: The ability to manage alerts for all New Relic products
- Full-stack monitoring: The flexibility to alert on what matters to your area of focus and responsibility, at any or all parts of the stack
- Manage incidents, not violations: The value of aggregating alerts to increase context and reduce noise
- Extensibility through integrations: The importance of connecting New Relic Alerts to whatever tools you already use for notifications and incident lifecycle management
Again, we want to give our customers powerful and flexible tools to monitor up and down the stack, and to help them be proactively notified when software enabling the business isn’t performant or healthy. So we’re not stopping there. Extending the themes of flexibility and customization, we are also excited to announce a preview of new alerting capabilities in limited release!
New Alerting feature previews
Dynamic targeting for dynamic infrastructure: New technologies such as cloud infrastructure and containers enable teams to easily scale resources up or down (often automatically) to deal with unexpected load changes or failure scenarios. That’s great for maintaining a positive customer experience, but challenging to monitor and accurately alert when things are coming and going all the time. New Relic Alerts is now integrated with our tagging system to automatically apply existing alert conditions and policies to changing instances of applications and infrastructure, so that teams can be assured that environmental changes are included in alert policies without requiring manual configuration.
Alerting + NRQL = magic: Many of our customers already know and use the power of New Relic Insights by writing NRQL queries to see their data in the ways that work best for them. We’re bringing the power of NRQL and Insights directly into New Relic Alerts by hooking it up to NRDB, the data store behind Insights. This enables teams to craft highly customized and precise queries that provide incredibly powerful signal data, which can then be alerted on. The possibilities are virtually endless, including alerting on customer experience or business metrics, or transactions that generate a specific response code.
We’ve also created a curated experience specific to alerting on percentiles, since percentiles are often better than averages because they give more weight to outliers (think SLA monitoring).
Powerful JVM alerting for Java apps: Java Virtual Machines (JVMs) are a core component of Java applications, and it’s not uncommon to have multiple JVMs supporting a single Java app. It’s important for teams to create alerts at this level to more easily troubleshoot issues with Java instances. So we’re announcing new capabilities that enable teams to create conditions for the JVM metrics they care most about, such as threads, garbage collection, and heap usage. But wait, there’s more! In addition to JVM-specific metrics, New Relic Alerts also now supports alerting on JVM instances in addition to the Java app itself. This helps teams proactively identify application performance issues with individual JVMs before they become bigger issues that impact customers.
Dynamic Baseline Alerts: We believe alerting best practices include setting thresholds for applications and services based on knowing the profile of expected behavior. But on-call operations teams, for example, are often responsible for applications they didn’t write and of which they may have only limited understanding. This is why we’re really excited to announce a preview of Dynamic Baseline Alerts, which leverages historical data to generate a baseline signal to set up dynamic thresholds for deviations of that baseline. We’ve worked hard to create a unique UI experience that will tell you when you would have been alerted based on the thresholds you set. Now you can create alerts with confidence! (For more information, see Introducing New Relic’s Dynamic Baseline Alerts.)
So now that we’ve cut the tape on New Relic Alerts, if you haven’t already, please go try it out for yourself. And if you’re attending FutureStack16 in San Francisco this week, come visit the Alerts team in person for a hands-on preview of the awesome new features.