Files
computer-playbook/tests/unit/python
kevinveenbirkenbachandClaude Opus 5.5 7b9ceb0bef fix(web-app-discourse): size the web workers by the host's vCPUs and RAM
UNICORN_WORKERS was hard-coded to 8. On a 4 vCPU / 16 GB CI runner that holds about 4 GB of web-worker RSS beside every other application of the round; the Nextcloud#7 rescue snapshot of fork run 42305502796 shows 8 workers of about 500 MB and 6.8 GiB of swap in use. The count is now clamp(min(2 * vCPU, RAM // 4 GiB), 2, 8): upstream's setup-wizard rule of two per vCPU up to 8, capped at one worker per 4 GiB of host RAM because the host is shared. The CI runner gets 3; hosts with at least 4 vCPU and 32 GiB keep 8. Only the env changes, so the launcher recreates the container without a rebuild (03_launcher.yml:50 drops env from the bootstrap digest).

🎯 Fix: 92 %
💥 Break: 4 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 88
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-07 15:05:49 +00:00
..