Architecture

Acceldata is composed of a set of services which are deployed using Docker. We host the production images on our docker repository hosted on Elastic container registry

architecture

Platform support

We designed Acceldata to be platform agnostic, we support all major distributions using a custom built abstraction layer and integrating with the underlying cluster services.

Acceldata modules

Acceldata services are composed of many modules

  • Base software: Includes basic connectors, the Acceldata database and Time series DB
  • Authentication: Authentication service and LDAP connector
  • Alerts: The alert engine
  • Log-search: The log collector and indexing service
  • Kafka connector: Kafka metrics collector

Your license terms allow you to run a subset or all of these services, the CLI will allow you to select which modules you can activate during the startup process

cli_select

Alerts Engine Architecture

Alerting engine is composed of many sub-systems.

  • It interally has a scheduler which triggers an evaluation of an alert.
  • An evaluator looks for the alert's definition and fires queries to the time series database and gives the result to a meter.
  • Meter takes care of accumulating evaluation results and firing them to the notification router when a threshold crosses.
  • The notification router then routes the message to one or more notification system based on alert definition.

alerts_overview