feat: add support for Alibaba Cloud AnalyticDB for PostgreSQL Nova

[softhuafei]

Summary

Adds a new client adbpgnova for Alibaba Cloud AnalyticDB for PostgreSQL (ADB-PG), which supports the Nova vector index
family (novamr / novad) backed by the fastann engine and USING ann syntax.

What's added

• Client implementation: vectordb_bench/backend/clients/adbpg/{adbpg.py,config.py,cli.py}
• CLI command: vectordbbench adbpgnova ...
• Frontend integration: db case configs + style entry, so the new client shows up in the Streamlit UI
• Unit tests: tests/test_adbpg.py — 9 tests covering config defaults, to_dict connection options, index_param WITH-clause options (incl. raw auto_reduction=on), session_param
  GUC emission, and TestResult.read_file round-trip
• README: a new "Run ADBPG (Aliyun AnalyticDB for PostgreSQL) from command line" section with examples

Algorithms supported

Algorithm	Build params	Search GUCs
novamr (default)	hnsw_m, ef_construction, rabitq_bits, auto_reduction	fastann.hnsw_ef_search, fastann.hnsw_max_scan_points, fastann.quantize_rescore_amp, fastann.nova_adaptive_gamma
novad	nlist, rabitq_bits	fastann.nova_nprobe (exposed as --nprobe)

Example

vectordbbench adbpgnova --case-type Performance768D1M --k 10 \
  --host <host> --port 5432 --db-name postgres \
  --user-name <user> --password <pwd> \
  --algorithm novamr --hnsw-m 48 --ef-construction 600 \
  --ef-search 50 --max-scan-points 16000 --quantize-rescore-amp 0.6 \
  --auto-reduction --num-concurrency 48

Or via YAML:

vectordbbench adbpgnova --config-file adbpg_cohere1m_novamr.yml

Test plan

• make lint (black + ruff) passes locally
• pytest tests/test_adbpg.py — 9/9 passing
• End-to-end runs against a live ADB-PG cluster for both algorithms (no-filter and 1%-filter cases)
• Streamlit Results page renders the saved JSON correctly

[softhuafei]

/assign @XuanYang-cn

[softhuafei]

can you please review @sre-ci-robot @jkatz @javiervegas @claude

[swhhy]

Hi @XuanYang-cn, this PR is ready for review. All CI checks (lint + unittest) pass locally. Would appreciate your review when you have a moment. Thanks!
/assign @XuanYang-cn 😀

[XuanYang-cn]

@softhuafei Please also resolve the conflicts.

[softhuafei]

@softhuafei Please also resolve the conflicts.
Hi @XuanYang-cn, thanks for the review. Addressed both blocking items:

Algorithm dropdown — vectordb_bench/frontend/config/dbCaseConfigs.py: options were ["novamr", "novam"], fixed to ["novamr", "novad"]. The backend (adbpg/config.py) only supports novamr and novad; novam was a typo that would let the UI submit an unsupported algorithm. Also fixed a related index-name comment in adbpg.py (novam_index → novamr_index).
Merge conflicts — rebased onto the latest main (linear history). The only conflict was in CASE_CONFIG_MAP, where both this branch and main appended an entry; both DB.Adbpg and the new DB.TurboPuffer are kept.
Could you take another look? Thanks!

[sre-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: softhuafei, XuanYang-cn

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [XuanYang-cn]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment