[Enhancement] Rate limit in order to not crash DefectDojo pods

[afagund]

Description

We are currently running the trivy-dojo-report-operator across multiple clusters, which is placing a significant load on the DefectDojo database. This increased load is causing crashes, unavailability, and noticeable slowness when accessing the DefectDojo console.

As an interim solution, we have implemented rate limiting at the Ingress level for DefectDojo. However, we would like to request a built-in feature in the operator to help mitigate these issues more effectively and prevent performance degradation.

Additional information

No response

[lbi22]

Hey @afagund @FieteO, I tested a change that should solve the issue. Could you try to use this in your chart to see if it also solve your issue?

image:
  repository: ghcr.io/lbi22/trivy-dojo-report-operator
  tag: test-latest

[manuel-sommer]

How exactly are you using trivy operator in prod? I am looking into how to use it productively, but it seems like trivy operator constantly reimporting the findings, spamming DefectDojo.

[quirinziessler]

This approach does not really help as long there is no mechanism to close old findings. Sending in batches is a good idea but that way DD is not able to check if existing findings are still valid as always another batch is being send and the logic of DD does only accept "full" imports for the auto-close function.

If you add a check for existing findings in DD and only push the new findings it would work. Maybe also sending batches is then not required.

[manuel-sommer]

#142