Fixes job 26203 (admin:test on develop): two tests that wait on async UI hit jest's 5s timeout.
-
funds-section.test.tsx› shows the error and retries -
order-legs-panel.test.tsx› lifts a balance refusal…
Root cause: admin:test had no --maxWorkers. jest sized its worker pool from the host core count and started 11 jsdom workers (holapc reports 12 CPUs), but the job is capped at 4 CPUs. The job ran alongside be:test and took 812s; it usually takes 48–220s. Both tests pass in under 1s on their own.
Fix: pnpm test --ci --maxWorkers=4, the same cap be:test uses. The tests are unchanged.
Verification: full admin suite passes locally, 828/828 in 21.7s. --showConfig shows maxWorkers: 4. This MR pipeline's admin:test should come back in its normal time range.
Review: Opus gate PASS in round 1 with 2 NITs, both kept as-is:
- The comment wording matches
be.yml. - Workers stay at 4 rather than 3 to match
be:test.