Readiness
- Readiness
- Production Ready
- Smoke
- passed
- Artifacts
- passed
- Production
- ready
- Storage
- <5
- Strategy
- Production Ready: real molecular-descriptors runner adapter deployed to Ubuntu backend; RDKit-core descriptor mode validated through runner API and Slurm.
- Next
- Production Ready for RDKit-core descriptors. Optional hardening: validate descriptor_set=mordred through MORDRED_PYTHON and add richer UI controls for descriptor set selection.
No obvious large model/database dependency identified; validate upstream docs before install. 2026-07-05 SOP batch recheck: backend inventory found no staged molecular-descriptors runner/artifacts, but existing RDKit/Mordred backend paths are production-proven for other tools. Kept Smoke Pending pending direct Slurm smoke and runner wiring. 2026-07-05 18:29 +08:00 backend re-audit: slug is not in runner registry and no object-store artifacts exist; existing mordred-molecular-descriptor-calculator runner is reusable evidence only, not promotion evidence for this slug. Kept Smoke Pending. 2026-07-06 implementation pass: ran direct Slurm job 443 at /media/nik/seagate_nik/bio_server/runs/molecular-descriptors/smoke_20260706_032000 using RDKit 2026.03.3. Aspirin and caffeine descriptor smoke produced finite MW, LogP, TPSA, HBD, HBA, rotatable bond, QED, and heavy-atom-count rows in results.csv plus summary.json, job.log, slurm.out, and slurm.err. 2026-07-07 17:40 +0800: SOP resolution pass wired a local real molecular-descriptors adapter in server/runner_registry.py, cluster_runner.py, and cluster_tools/cpu_tools.py. Adapter dispatches RDKit-core descriptors by default and Mordred when descriptor_set=mordred/all/full; it now fails if RDKit/Mordred is unavailable instead of producing proxy descriptors. Audit sees runner_status=real_runner_configured locally, but Production Ready is deferred until updated code is deployed on Ubuntu and validated through live API/Slurm artifact downloads. 2026-07-07 17:44 +0800: Promoted to Production Ready after deploying local adapter to /home/nik/bio_mega_runner on Ubuntu backend and restarting runner. Runner API job fa109c879105 submitted Slurm job 497 for aspirin RDKit-core descriptors and completed. Artifacts downloadable through runner manifest: summary.json, results.csv, job.log, slurm.out, slurm.err. Observed metrics: MW 180.159, LogP 1.3101, TPSA 63.6, HBD 1, HBA 3, rotatable_bonds 2, QED 0.5501, heavy_atom_count 13.