Note: We've updated our pricing and packaging. This feature is available only on the new plans.


Freshservice Alert Management enables ITOps & NOC Teams to oversee the status of the IT infrastructure issues from within their IT service desk. Freshservice Alert Management tackles noise, provides a real time view of alerts on a single pane of glass, and automates incident creation, routing, and resolution. Here’s a bird’s eye view of how it is done.


  • Notifications from all monitoring tools are piped into Freshservice using webhooks

  • The payload is standardized by mapping the values used by monitoring tools to the ones used within Freshservice

  • Noise is reduced by grouping notifications from a common resource and with a common metric

  • The swarm of signals sent by multiple monitoring tools is thus filtered to extract a list of alerts on a single pane of glass

  • Algorithmically defined conditions drive automated conversion of alerts into incidents, and their resolution

  • Such algorithmically defined conditions also automatically route incidents to agents and agent groups

  • Users with sufficient number of resources and alerts qualify for advanced noise reduction features using Freddy Machine Learning algorithm




Integration of monitoring tools in Freshservice

While Freshservice offers out-of-the-box integrations with a growing list of popular monitoring tools, it can integrate with any tool in the world using webhooks.


In either case, the payload is normalised and severity values mapped to Freshservice values. Establishing a common standard to interpret notifications from multiple monitoring tools has dual benefits:

  • Users are able to comprehend issues faster

  • Freshservice is able to detect correlation between notifications from different monitoring tools and with other varying metadata to identify underlying issues faster 


Noise reduction in Freshservice

Freshservice reduces noise in three ways to simplify alert management:

  • Algorithmic grouping: Notifications from monitoring tools are grouped together on the basis of a common resource and metric. This helps suppress noise from repetitive and duplicate alerts. In Freshservice, a group of such notifications is referred to as an alert. 

  • Machine Learning based Automated Grouping: Freddy, the Machine Learning algorithm, studies the historical patterns between alerts and associated incidents. Based on those patterns, Freddy attaches relevant incoming alerts to open incidents. Doing so preempts the unnecessary creation of a new incident while making an existing incident contextually richer. In Freshservice, we refer to this strategy as Automated Grouping. 

  • Manual association of an alert with an open incidentAn alert can also be manually attached to a relevant open incident. This not only reduces noise at the incident level, but also trains Freddy ML algorithm to learn from this behaviour in accounts for which it is enabled.  


Accessing Alerts within Freshservice

Once you’ve set up Alert Management, Freshservice will start receiving notifications from your monitoring tools. These notifications are treated for noise reduction after which alerts are displayed on the Alerts List page which can be accessed by clicking on the Alerts icon on the left nav bar. 




View all open alerts under the Alert ID column. Each alert has an associated subject, severity, last updated details, the resource, and the associated incident if any. 


Use the filter option on the top right corner to filter alerts based on created date, severity, alert profile, resource, etc. 



 

Incident creation and resolution

Freshservice enables automatic creation and resolution of Incidents based on Alert Rules. An Alert Rule is a set of conditions that an alert, generated by a monitoring tool, must meet to be treated as an incident. It also contains instructions for the automatic routing of an incident to the relevant.



Now that you are familiar with the concept of Alert Management in Freshservice, check out these related resources to start making the most of this feature: