Conditional attributes processor

[tamirdavid1]

No description provided.

[RonFed]

General questions:

1. Is the main goal here is to improve CPU consumption from the old approach?
2. Do we know how much CPU was consumed by the old processor (in terms of percentages)?
3. How is this change improving the CPU consumption relative to the current approach? Did we measure the difference?
4. Do we need to delete the old implementation? I didn't see that in the PR but maybe I missed that.

[tamirdavid1]

General questions:

1. Is the main goal here is to improve CPU consumption from the old approach?
2. Do we know how much CPU was consumed by the old processor (in terms of percentages)?
3. How is this change improving the CPU consumption relative to the current approach? Did we measure the difference?
4. Do we need to delete the old implementation? I didn't see that in the PR but maybe I missed that.

Before starting responding, this is the first phase of this change the next phase will include creating an action that will create this processor with needed configuration. after validation I will delete the old approach.

Link to the old approach of setting category attributes:
https://github.com/odigos-io/odigos/blob/main/cli/cmd/resources/profiles/category-attributes.yaml

Hey @RonFed ,

1. Yes the main goal is to improve the CPU compared to the old approach (transform processor)
2. We dont know how much CPU was consumed but it was high cpu consumer, imagine that for each span we went over the huge list of the category attribute that include also regex
3. In the suggested approach for each condition we can ask, what are the new attributes values based on another attribute with value X.
   It uses map so dont need to iterate just ask values[odigos.io/ebpf/instrumentation-fastify]
   
   
   For your question we didnt measure the improvement yet, will do so in the next phase of this change.
4. after i'll make sure that it works i will remove the old approach