The GitHub Actions job "License Binary Checker" on texera.git/main has failed. Run started by GitHub user github-merge-queue[bot] (triggered by github-merge-queue[bot]).
Head commit for run: bcd0ec4f742264503114aa4388d82df8d35e9647 / Meng Wang <[email protected]> fix(deploy): load example datasets against the account's real email (#8723) ### What changes were proposed in this PR? `bin/single-node/examples/load-examples.sh` derived the owner email that file-service matches dataset lookups against by concatenating `@example.com` onto the username. On a single-node deployment that never matches: the admin seeded from `USER_SYS_ADMIN_USERNAME` has the username itself as its email (`AuthResource.createAdminUser`), so the loader logged in as that admin, skipped its register branch, and every example file upload failed with `400 Dataset not found`. Both example datasets were created but left with no files and no version, so the two example workflows pointed at CSV paths that do not exist. The owner email now comes from the access token's `email` claim, which is whatever the server actually stored — correct for the seeded admin and for an account the script had to register itself. The register call keeps an `@`-shaped address, which the server's format check requires. Second, the loader now exits non-zero once it has logged an error. Each failure is still non-fatal so one bad file does not abandon the rest, but they used to leave no trace: the container printed `=== Example data loading complete ===` and exited 0. That is why the broken lookup went unnoticed for two months. ### Any related issues, documentation, discussions? Closes #8721. Introduced by #6402, which had to give the register call a regex-valid address and changed the shared variable rather than splitting its two roles. `release/v1.3` carries it too. ### How was this PR tested? New `bin/single-node/tests/test_load_examples_sh.sh` (9 assertions, picked up automatically by build.yml's shell-test discovery) drives the loader against a stub `curl` — no docker, no network, no sleeping. It pins the owner email to the token's claim, asserts it is never `<username>@example.com`, and asserts that a failed upload or version create exits non-zero. Both fixes were reverted individually to confirm the relevant assertions go red and the suite exits 1. End to end on a fresh single-node deployment (`down --volumes` then `up --with-examples`): the loader now reports `Owner email resolved as 'texera'` and uploads both files, `dataset_version` goes from 0 rows to 2 (both `v1`), and presigned download of the two paths the example workflows scan returns HTTP 200. ### Was this PR authored or co-authored using generative AI tooling? Generated-by: Claude Code (claude-opus-5) Report URL: https://github.com/apache/texera/actions/runs/36413531226 With regards, GitHub Actions via GitBox
