Complex Event Processing (CEP)
Prevent recurring problems with Real-time analysis
Improving signal-to-noise ratio
The advanced application analysis engine using Complex Event Processing is the core of AutoPilot. It performs event stream processing and correlation. The analytical engine can be distributed, collaborate and linked together using high speed pub/sub – and capable of almost linear scalability in accordance with load.
Analytics for Complex Event Processing (CEP) Features include:
Real-time rules processing
Time-based trend recognition
Aggregation, sorting, merging and joining of events and metric
Scales to handle millions of events per second
Analytics (mathematical computations such as Standard deviation, EMA (Exponentially moving avg.), MA (moving avg.), momentum indicators and many more) to automatically determine business normal vs. abnormal states
Wizard driven GUI interface
Predictive problem prevention capabilities
Preventing recurring problems
AutoPilot’s real-time application analytics engine using complex event processing (CEP) automatically calculates dynamic trends, creating its own metrics such as “change in rate-of-change” and effectively determine what is truly outside of business normal for your environment. It can instantly differentiate spikes and resource consumption from a true business impacting problem that might cause an outage.
These metrics can be composite metrics or KPIs that are based on multiple conditions specified by business policies and are used for:
Compliance – Detect potential or actual breaches in responsibility
Prevent false alarms
Understand trends across composite applications
Determine if you will be able to handle rapid increases in load
Keep up with the big-data from events and ensure that authentic problems are attended to
Elastic Applications require Elastic Monitoring
Today’s applications are elastic, either deployed in virtual machines, clusters or in the cloud. Elastic applications can’t be monitored using yesterday’s static models. Static thresholds and modeled transaction flows become outdated almost instantly and are ineffective. Constant change is the new normal and this makes root-cause analysis very difficult. An elastic approach with real-time monitoring and analytics, which creates automatic dynamic trends across domains, is essential to reduce the frequency and duration of incidents and detect problems before there is impact.