Bringing on-call and incident information into Compass

[#1]

Hey Community, our team is super excited to share that Compass now has Opsgenie & PagerDuty apps AND there’s a new kid on the block in the Activity Feed - Incident Events! Oh and while we were at it, we tossed in a bonus feature: each unique event source will now be given its own row on the activity feed so you know where the events are coming from  

On-call information

The Opsgenie and PagerDuty apps bring your on-call schedules directly into the component detail view so you can find the right people at the right time when troubleshooting or navigating an active incident. Simply drop a link to your Opsgenie schedule or PagerDuty service into the on-call section of your component.

Incident events in the activity feed

Compass can now ingest and plot incidents in the activity feed alongside deployments, and other events, for your components and their dependencies. First install/configure either integration and then follow the documentation (Opsgenie, PagerDuty) to get incident events flowing into compass. You can also learn more about sending events, including incidents, to Compass using our REST API.

Quick callout: Our Opsgenie integration is an early version that we wanted to get in your hands while we continue to improve upon it. You may encounter some rough edges getting incident info flowing in, check the docs for setup info and callouts.

Incident metrics

After incident events start flowing in, the new apps also give you key metrics like “Incident Count” and “Mean Time to Resolution" so you can better understand the health of your components. Whenever an incident is opened or closed these metrics will be automatically calculated for you; there is no configuration required. And great news, you can also use these metrics with scorecards to track these indicators across collections of components.

Checkout the apps today

Compass provides deeper visibility into how incidents impact all software components in your distributed architecture while giving you a DevOps health platform to prevent future incidents. Checkout the Opsgenie and PagerDuty apps today by navigating to the “Apps” menu in Compass, and let us know what you think using the “Give feedback” button.

Happy Compassing!

[Davidovitz]

Hi, This is a fantastic integration and I'm looking for more OOB metrics !!!

While trying these metrics in Score card I saw this behavior :

If you create a ScoreCard on “ Incident Count” for example:   Incident Count < 2 , the condition will return false while incident count = None  ( no incidents) . I expect it to return true since 0 ,  but behave like Null.

[Josh Campbell]

Hey Davidovitz! There's some nuance behind the scenes that makes this an interesting one :) As far as scorecards go, this is actually working as intended. If a metric value is null for a component it will be considered as failing the scorecard criteria; it can't be passing with no data. That said, we do plan to ship some UX improvements soon that will help explain various scorecard states (e.g. you are missing data for this metric so it is failing).

On the incident count metric end, we need to backfill the data to grab historical values (before connecting to component) so we're seeding your components with good data. We don't currently backfill incidents which is why it comes in as null even if you had incidents recently (however, a new incident for that component should trigger some calculations to happen). I don't have an ETA for the backfill to be implemented but it's on the radar.