lhumina_code/home
22
0
Fork 0
Code Issues 3 Pull requests Projects Releases Packages Wiki Activity Actions

E2E tests: 54 WASM shell timeouts with native 15MB build #106

New issue
Open
opened 2026-03-31 19:53:34 +00:00 by mik-tf · 0 comments
mik-tf commented 2026-03-31 19:53:34 +00:00
Owner
Copy link

Problem

With the native Dioxus build (web-native feature, 15MB WASM), 54 out of 167 Playwright E2E tests fail with 5-minute timeouts. All failures are tests that require full WASM shell navigation (login → store → open island). Tests that use direct HTTP/API calls pass fine.

Root cause

The 15MB native WASM bundle takes 60-120s to load in headless Chromium. When Playwright runs multiple browser contexts in parallel, they compete for CPU/memory, pushing load times beyond the 5-minute timeout.

In session 6, the same tests passed (40/41) because fewer tests ran or had less resource contention.

Test results (session 7)

Layer Result
Smoke 122/124 (2 skipped)
Integration 18/19 (1 fail — Redis API change, unrelated)
Playwright E2E 111/167 (54 timeouts, 1 real fail, 1 edge case)

Non-timeout failures

  • Compute CP6: RPC discover — instant fail (12ms), likely API change from development merge
  • Conversation persistence — 1.3m, WASM load edge case timeout

Options to fix

  1. Increase timeout to 600s (10 min) — simplest, doubles test runtime
  2. Limit Playwright workers to 1 (workers: 1 in config) — serializes tests, each gets dedicated resources
  3. Split test suites — run API-only tests separately from shell-navigation tests
  4. Combination — workers=1 for shell tests, parallel for API tests

Related

  • Part of #104 (Dioxus SPA migration)
  • Iframe build (9.5MB WASM) does not have this problem
## Problem With the native Dioxus build (`web-native` feature, 15MB WASM), 54 out of 167 Playwright E2E tests fail with 5-minute timeouts. All failures are tests that require full WASM shell navigation (login → store → open island). Tests that use direct HTTP/API calls pass fine. ## Root cause The 15MB native WASM bundle takes 60-120s to load in headless Chromium. When Playwright runs multiple browser contexts in parallel, they compete for CPU/memory, pushing load times beyond the 5-minute timeout. In session 6, the same tests passed (40/41) because fewer tests ran or had less resource contention. ## Test results (session 7) | Layer | Result | |-------|--------| | Smoke | 122/124 (2 skipped) | | Integration | 18/19 (1 fail — Redis API change, unrelated) | | Playwright E2E | 111/167 (54 timeouts, 1 real fail, 1 edge case) | ## Non-timeout failures - **Compute CP6: RPC discover** — instant fail (12ms), likely API change from development merge - **Conversation persistence** — 1.3m, WASM load edge case timeout ## Options to fix 1. **Increase timeout to 600s** (10 min) — simplest, doubles test runtime 2. **Limit Playwright workers to 1** (`workers: 1` in config) — serializes tests, each gets dedicated resources 3. **Split test suites** — run API-only tests separately from shell-navigation tests 4. **Combination** — workers=1 for shell tests, parallel for API tests ## Related - Part of https://forge.ourworld.tf/lhumina_code/home/issues/104 (Dioxus SPA migration) - Iframe build (9.5MB WASM) does not have this problem
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
development
main
No results found.
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
lhumina_code/home#106
No description provided.