Context:

With OPT-5953, we have added a basic zeebe data generator job to generate data we can use in import performance tests. We can iterate on this generator to improve it by:

• ability to create different data snapshots, eg one 1M instance one for import performance tests and one 100M instance dataset for query performance tests
  • ability to select name/filter of snapshot as job parameter done
• consider adding manipulation of when instances are exported to create zeebe record indices spread over several days
• consider adding the job worker count as a configurable parameter done
• consider adding more variety to deployed instances/definitions
• extract methods shared between zeebe data generation and import performance into shared jenkins lib (eg ES snapshot related code)
• switch to stable node pool to avoid issues with longer running jobs done

AT:

• tbd depending on which of the above are to be implemented, might also extract tasks into separate tickets