Fix sticking of Runner a new job is scheduled

What does this MR do?

Fixes Runner stickiness when a new job is scheduled and we have significant replication lag.

Instead of saying to runner: "Hi, nothing for you, go away" we say "Hi, something is wrong, try again later".

This will make jobs/request to continue using replicas, always, but gracefully react for replication lag, and retry.

Does this MR meet the acceptance criteria?

• Changelog entry added, if necessary
• Tests added for this feature/bug
• Conform by the code review guidelines
  • Has been reviewed by a Backend maintainer
  • Has been reviewed by a Database specialist
• EE specific content should be in the top level /ee folder
• Conform by the merge request performance guides
• Conform by the style guides
• If you have multiple commits, please combine them into a few logically organized commits by squashing them
• Internationalization required/considered
• If paid feature, have we considered GitLab.com plan and how it works for groups and is there a design for promoting it to users who aren't on the correct plan
• End-to-end tests pass (package-qa manual pipeline job)

What are the relevant issue numbers?

Related to https://gitlab.com/gitlab-org/gitlab-ce/issues/46037