Goodbye, Incidents. Hello, Alerts.

Posted by Owen Zanzal on Jun 6, 2016 7:14:50 PM

All VividCortex users, please repeat: “Goodbye, incidents. Hello, alerts.”

We’re changing the way that VividCortex tells you when you need to pay attention to a specific occurrence in your system. We recently phased out VividCortex’s “incidents” services, and the “incidents” section in our app’s UI is gone. This isn’t just a cosmetic change. We’re migrating the incident configurations into a whole new concept: alerts.

alarm_clock.jpg

Image Credit: By Sun Ladder (Own work) CC BY-SA 3.0, via Wikimedia Commons

VividCortex alerts are currently in beta testing. They allow you and your team to be notified when important criteria are met in your environment. Specifically, alerts can be calibrated to generate when an Event matches a specific set of filters or when a Metric exceeds defined threshold parameters. As such, there are two different types of alerts: Event alerts and Threshold alerts. As the names imply, Event alerts generate whenever a designated event happens in your account. And, naturally, Threshold alerts generate whenever a metric exceeds a designated threshold. 

Integration is supported too -- when an alert is triggered, the resulting message can be routed to one or more Integration Destinations. This means you can set up the notification so that it appears in a variety of places, based on your preference and convenience. Current Integration Destinations include email, Hipchat, PagerDuty, Slack, and VictorOps. With these integrations intelligently applied, you should never miss an important alert.

One of the major changes between incidents and alerts is that incidents required a customer to manually close each incident, whereas alerts do not. We paid attention and listened to much customer feedback, where users expressed a preference for not needing to manually close each and every incident that fired. In fact, in actual practice, many customers ultimately chose to just never close incidents—this non-use is the sure sign of a sub-optimal feature. We’ve listened, and we’ve designed alerts so that, unlike incidents, they don’t require manual acknowledgement. Instead, users can designate a duration window during which time an alert cannot repeat. However, the alert is allowed to trigger again once the window resets. This keeps flickering alerts from blowing up your system, plus it allows users to customize the duration window based on exactly what an alert is looking for and how urgently the user would like to respond whenever it goes off.

On top of all this, we wanted to move away from “incidents” because VividCortex is not specifically designed to be an incident management app. Other products, such as PagerDuty and VictorOps, are, and VividCortex is at its best when used in conjunction with services like those. Customers who want incident management can achieve it by using a management app (like VictorOps) and setting the alert duration window to a short amount of time.

The bottomline is that with alerts you’ll find more customization along with more automation — a powerful combo. Goodbye, database clutter. Hello, VividCortex.

Let us know what you think!

Recent Posts

Posts by Topic

see all