Retry Intervals

I am seeing some behavior on specifying Retry Interval, which I seeks some advice on.
Retry Interval is defined as “PT10S, PT20S, PT40S”. However the retries happened like this

21:29:09.946
21:29:25.002 ==> 15s from last occurrence
21:30:00.043 ==> 35s from last occurrence
21:31:15.091 ==> 75s from last occurrence

Is the retry interval not guaranteed? Or is it only on best effort basis?

I am using the following stack:
camunda-bpm-spring-boot-starter:7.13.0
h2 in memory database for my POC
spring-boot-starter-parent: 2.6.0

All of the “time” based elements/configuration are best effort based, as you said. All this will do is to configure the job to NOT EXECUTE until this time, updating the act_ru_job Due Date column. After this due date time passes, it will be free to any job executor take this job and execute.

This will be a work of tunning your environment to the size of your job queuee, putting more containers/job executors, adjusting the quantity of jobs per job acquisition, job executor threads, and your job acquisition polling intervals. All of this configuration can be done in spring properties.

I really love this page about camunda performance tunning, and maybe it will help you find your way :slight_smile:
https://camunda.com/best-practices/performance-tuning-camunda