trivy multi-replcas functioning

[rammanoj]

I am running 3 replicas of trivy operator on my system. I've deployed a huge load of containers on the system and trivy started and completed the scans successfully 🎉

However, when I observe the resource that these three pods consume, one of the three is consuming a lot more resources when compared to the other 2. I am currently running with scanJobsConcurrentLimit set to 10. PFA RAM snapshot.

My assumption is that the load would be distributed across all three pods in a similar fashion, isn't it the scenario? Can I get some context on how is this working?

[chen-keinan]

@rammanoj can you share the consuming operator log

[rammanoj]

@chen-keinan, just curious to know if you find something regarding the above.

[chen-keinan]

@rammanoj is the operator reconcile all resources? have you configure the leader election ?

[rammanoj]

@chen-keinan I have not configured the leader election explicitly. Doesn't it gets auto-enabled? I could not find any other option on the values.yaml file to enable the election explicitly. The only option I could see regarding the election is leaderElectionId. Is there any way through which I could enable it?

[chen-keinan]

Actually it should happen automatically when replica is gt than 1

[rammanoj]

cool, I have not tampered with its behavior. So, is it consuming more resources because it's acting as a leader and reconciling the resources?

[chen-keinan]

only the leader should perform the reconcile

[rammanoj]

So, is this pod consuming more resources because it could a leader?

[chen-keinan]

I think so , need to be checked.
also easy to test , scale down to 1 and check it resources