ML-based situations

A  situation comprises events associated with the same or different hosts that are aggregated based on their occurrence, message, topology, or a combination of these factors. Events are collected from multiple sources across infrastructure, application, and network resources available from various monitoring solution vendors.

As a tenant administrator or a custom user with manage situations permissions, you can create an ML-based event aggregation to:

  • Derive actionable insights.
  • Investigate the aggregated events.
  • Reduce the event noise.
  • Improve the mean time to resolve (MTTR) based on the situation driven workflow.
  • Lower the mean time to detect or discover (MTTD) and the time required for investigating tickets.

AIOps or ML-based situations

The ML-based situation uses an AI/ML-based event processing technique. The algorithm identifies event patterns from hundreds of raw events, filter out those events, and automatically groups those events into situations with a unique signature added to each event type. The signature could be different for each issue type of the situational event. The AI/ML algorithm uses the event signature to build unique situation from the aggregated events. 

For example, events from a node or business service in BMC Helix Discovery are processed and each event is assigned a unique signature. For example, a CPU utilization event signature is different from a memory utilization event of the node. 

Event signature

An event signature is a pattern used to categorize an event from an event sequence as belonging to either same or different category. However, the temporal relationship between these events are also considered to aggregate these events as part of a situation. For example, an event sequence that has CPU utilization at 90%, 93%, and 97% all have the same signature provided they are all within a defined time window to aggregate them as part of a situation.

Prerequisites to viewing ML-based situations

For creating and viewing ML-based situations in BMC Helix AIOps, learn how to:

  • Create a service model or business service in BMC Helix Discovery
  • Configure BMC Helix Operations Management to process events
  • Enable the AIOps features in BMC Helix AIOps

Benefits of building ML-based situations

With the ML-based situations, you get the following benefits:

  • View highly-contextualized situations without having to create policies or rules to aggregate these events.
  • No prior-knowledge about what rules need to be part of the policies is required.
  • Ability to go through tons of such situations without having to individually go through all of them, as the algorithm analyzes the topological and temporal relationships to group these events.
  • Forms the basis of root cause analysis and isolation of causal nodes.

Illustration

The following diagram shows how the situations are created from the raw events:

Was this page helpful? Yes No Submitting... Thank you

Comments