Using SQL as the Key to Robust Automation
Kelverion Best Practices; Using a Persistent Data Store
It is Kelverion’s recommendation that in nearly every case runbook automations be implemented utilizing a small backend SQL database table or set of tables. This document explains why we make this recommendation.
The ORCHESTRATOR data bus is extremely powerful but it is not persistent. This means that if the ORCHESTRATOR runbook server were to reboot or the services were to crash all data about executing runbooks would be lost. So a complicated multistep workflow that is counting on the data bus for its current state would need to be restarted manually by an administrator. That administrator would potentially need to take significant time to understand what parts of the workflow had run successfully and what parts were still needed. If a persistent data store is used at the backend, the last successful step in the process has been recorded in the database with all necessary information to continue the next step.
No material found.