[ehb54/us3lims_common] 651d22: Reset $output between sbatch retry attempts in att...
Branch: refs/heads/main Home: https://github.com/ehb54/us3lims_common Commit: 651d220f49c593fb9a5d43d699e8ac1797b0818b https://github.com/ehb54/us3lims_common/commit/651d220f49c593fb9a5d43d699e8a... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-06-24 (Wed, 24 Jun 2026) Changed paths: M class/submit_local.php Log Message: ----------- Reset $output between sbatch retry attempts in attemptSubmit() PHP's exec() appends to $output rather than replacing it. Since $output was only initialized once before the retry loop added in PR #21, parseSubmitResult()'s check of $output[0] (used to extract the slurm job ID) stayed permanently stuck on the first attempt's output for the rest of the retries, even after a later attempt genuinely succeeded. Confirmed live via us3lims_dbutils' uslims_autoflow_test.php --fake-sbatch fail-once: after a deliberately-failed first attempt, attempts 2-4 each succeeded for real (three separate, duplicate jobs submitted to the actual cluster) but were still misjudged as failures, so the whole submission ended up reported FAILED with one of the duplicate real jobs left running with no LIMS record of it. Fixes ehb54/ultrascan-tickets#931 Commit: 2a60c1339566ae809b747a57f9810040eaa2b306 https://github.com/ehb54/us3lims_common/commit/2a60c1339566ae809b747a57f9810... Author: emre brookes <ehb54@users.noreply.github.com> Date: 2026-06-24 (Wed, 24 Jun 2026) Changed paths: M class/submit_local.php Log Message: ----------- Merge pull request #22 from ehb54/ehb54-issue-931 Reset $output between sbatch retry attempts in attemptSubmit() Compare: https://github.com/ehb54/us3lims_common/compare/367f2e89e377...2a60c1339566 To unsubscribe from these emails, change your notification settings at https://github.com/ehb54/us3lims_common/settings/notifications
participants (1)
-
emre brookes