Change trigger logic of operator

[MathieuCesbron]

Hello,

At the moment, we reconcile a resource and then re-reconcile it forever every 10 seconds.

Advantage:

• If we delete by mistake a service link to the pods then the service will be recreated in 10 seconds in the worst case scenario.

Disadvantages:

• Pollution of logs
• Not elegant when there is lots of resources to reconcile since every resource will trigger the reconcile every 10s.

Do you think it would be better to not reconcile every 10s ? I would find it more elegant but I am waiting for other inputs.

Have a good day,

[shubham-cmyk]

@MathieuCesbron Can you propose a solution

[drivebyer]

Maybe we should watch the related resource, if any changes on it, then we reconcile it.

[shubham-cmyk]

@drivebyer We could store a deep copy of CR and then match it. if we find any changes we could reconcile if not just skip the reconcile.

this could be made optional by an annotation that could skip this reconcile