Rework total calculation for repository tree finder

Everyone can contribute. Help move this issue forward while earning points, leveling up and collecting rewards.

• Close this issue

The following discussion from !67509 (merged) should be addressed:

• @mkaeppler started a discussion: (+4 comments)

  Could you elaborate on this, i.e. inefficient how? Is this in reference to fetching all objects just to count them? Do we expect this to become a problem in production?

  Let's create a follow-up issue as well and link to it here, or it's unlikely anyone will actually replace it.

  A few questions about the implementation overall:

  • Why are caching the tree only when counting the total, when in execute we fetch it again anyway? Can we fetch-and-cache once, then count the item total?
  • Why are we fetching this into redis? I assume the total is rendered out together with the current page. Can we cache this in-process instead?
  • If we decide to use Redis, can we estimate how this will contribute to Redis memory growth over time?