When an anomaly is detected, Unryo performs root cause analysis (to determine the probable cause) and impact analysis (to identify the impacted resources).
Root Cause Analysis#
Root cause analysis is based on the topology knowledge. When a resource goes down, the Unryo Correlation Engine determines if the resource is the root-cause or an impact, according to the status of the parent resources. The root cause status is specified in the alert property, to facilitate triage and to reduce alert noise.
The Business Impact Analysis lets you define Business Services ; such as Customers, Department, Contracts and more ; then calculates their status, based on impact analysis driven by propagation rules.
How it works#
To do this, the
Analytics Engine uses an
Impact Tree to store all the relationships between the defined Business Services and the infrastructure components, such as hosts, network interfaces, server processes and more.
Analytics Engine computes in real-time the state for every component in the
Impact Tree. When an alert is received, the algorithm calculates the impact and determine how Business Services are affected. There is no limitation on the tree structure: you can configure as many levels you want, such as a business service (i.e. Customer) composed of other business services (i.e. Departments) with each of them composed of various infrastructure components (i.e. VMs, hosts, network devices, interfaces, processes, ...)
The impact propagation is defined by impact rules and weights (percentage, min up, max down), that are customizable.
Creating a Business Service Component#
Click on the
Correlation & Impact Analysis, then click
+ button to open the editor.
From there, you create your Business Service component, and customize the propagation rules:
|Select your Correlation Template||Choose
|Select your Correlation Engine||Indicate the
|Provide the new resource name||indicate a name that is meaningful for you, for example
|Resource Type||From the proposed list, pick a type that best describes your resource. You have the choice between
|This resource is considered ok if:||This setting lets you customize how the resource members will affect this Business Service, using a
|Monitor SLA||If you want SLA monitoring, specify the SLA % threshold objective. This will allow you later on to report on SLA and alert if the SLA objective is not met. Or choose
|Resource Members||Indicate one of multiple members for this Business Service. A member can be either a
Alerts on Business Services are visible in the Alerts Console:
Business Service dashboards show all defined services, with their status over time, their SLA and related alerts.