SonarCloud LOC limit inconsistently enforced - unsure if we actually need to upgrade our plan

SonarQube Cloud
1

Hi everyone,

We’re on a 100K LOC plan with 5 projects (around 250K LOC combined). Because of LOC limit errors, our usual workaround has been to delete or exclude a project before running analysis on the rest, to stay under the cap.

However, we’re seeing inconsistent enforcement:

  • Normally, analyzing projects that exceed 100K LOC combined triggers a “Lines of Code limit exceeded” error.
  • This time, 3 projects analyzed successfully together, with individual LOC of 99K, 98K, and 22K (219K total), well over our 100K cap, and no error was thrown.
  • On other occasions, we’ve hit the LOC error even when seemingly within range, forcing us to delete a project as a workaround.

Because of this inconsistency, we genuinely can’t tell if our current plan is sufficient or if we need to upgrade. If 219K LOC can pass without error, it’s unclear whether:
(a) our plan actually supports more than 100K LOC than advertised,
(b) LOC enforcement has a delay or sync issue that lets analysis through temporarily before blocking it later, or
(c) something else is going on that we’re misunderstanding about how LOC is counted.

Questions:

  1. Is LOC enforcement based on a periodic sync or snapshot rather than real time calculation? Could that explain why 219K LOC passed without error?
  2. Is there a delay between when analysis is submitted and when it counts toward the org’s LOC usage, meaning a run could succeed now and get blocked later even without new code changes?
  3. Is deleting or excluding a project a supported way to manage LOC limits, or could this cause side effects like stale LOC counts still being counted temporarily?
  4. Most importantly: given our actual usage is around 250K LOC, do we need to upgrade our plan to reliably run analysis on all 5 projects, or is there a way to work within our current 100K plan (for example using LOC exclusions for generated or vendor code)?

We would rather get clarity on this before deciding to upgrade, since right now we can’t tell if the errors are a hard limit being enforced late, or if there’s a genuine bug allowing us to exceed the cap.

Thanks!

2

Is deleting or excluding a project a supported way to manage LOC limits, or could this cause side effects like stale LOC counts still being counted temporarily?

No. Not the way you’re doing it. If you need to delete old projects that aren’t deployed anymore to free up loc for new projects – that’s supported. If you’re deleting projects on a rotating basis to avoid having to pay for additional loc – that isn’t supported.

Yes.