Brief summary of the bug. What is it ? Where is it ?
Context: https://camunda.slack.com/archives/C049FU181MJ/p1667505668542349
At least one of the management reports specifically targets process instance indexes. In cases where there are many processes deployed, this can cause the ES request to fail because the URI gets too long (too many indices targeted).
A simple solution would be to target the process instance index alias, but this might not be acceptable because users don't necessarily have authorization for all processes
An alternative would be to target the process instance index, but then to reduce the result size by using boolean conditions on the search. While this would solve the bug, it would result in worse performance than the current solution.
Note that it is important with either change above, or any alternative solution, that the preseeded customer onboarding data is excluded from the results.
Steps to reproduce:
Deploy a lot of processes (the number is unclear as it also depends on the name of the process)
Actual result:
The report evaluation fails
Expected result:
The report evaluation works with all the expected target processes
Bug Report
L3 - Default