test(e2e): raise Epic hierarchy indexing budget to 240s

What does this MR do and why?

The end-to-end indexing suite started failing on the default branch, timing out while waiting for freshly created entities to become queryable. Under post-merge load any of the suite's cases can exhaust its time budget before indexing catches up. This gives every case in that suite more headroom.

Relates to #926 (closed)

Testing

Per-case timeout bump in the E2E harness; verified by re-running the e2e job on this branch. A first run confirmed the project-issue-note case (not only the group-scoped epic) can also exhaust the old budget, so all three cases were raised.

Performance Analysis

Raises a test-only wait ceiling; no product code path changes.

  • This merge request does not introduce any performance regression. If a performance regression is expected, explain why.
Agent context

Started from !1962 (merged) (note_3514133340): the 02 Indexing Epic-hierarchy case exhausted its 180s budget on main once the E2E license was rotated back on (allow_failure: false). Failing job 15133264310.

A verification run on this branch (15134773024) then showed the first case, "Project Issue And Note Are Indexed" (the project path, not the epic), also exhausting the full budget on Note id=... not found. So the whole 02 suite is racing its budget under load, non-deterministically tipping on whichever case runs when Siphon/indexer throughput is lowest. Fix: raise Start Indexing Budget from 180 to 240 on all three cases in e2e/tests/02_indexing.robot (240 is already what case 06 uses).

That same verification run also hit transient 502 responses from the Orbit webservice in 05 and 08 (Expected status: 502 != 200). Those are environmental (webservice restart mid-run), absent from the original main failure, and unrelated to this budget change. Re-running to confirm the 02 suite goes green.

%{all_commits}

Edited by Michael Usachenko

Merge request reports

Loading
Loading