Merge article/03 lab into main: module 03 (10K corpus, IVF filter-first proof, HVI) + module 02/03 reconciliation - #8
Merged
Conversation
…vector-vs-converged Per-article companion module on its own branch (branch-per-piece), rehosted in the Oracle UMT Developer Hub layout under labs/converged-database-lab/.
|
Thank you for your pull request and welcome to our community! To contribute, please sign the Oracle Contributor Agreement (OCA). To sign the OCA, please create an Oracle account and sign the OCA in Oracle's Contributor Agreement Application. When signing the OCA, please provide your GitHub username. After signing the OCA and getting an OCA approval from Oracle, this PR will be automatically updated. If you are an Oracle employee, please make sure that you are a member of the main Oracle GitHub organization, and your membership in this organization is public. |
oracle-contributor-agreement
Bot
added
the
OCA Required
…r-first plan proof (import from working repo @bd039cc) - 10,000-ticket deterministic seed (tickets 1-300 byte-preserved; 301+ from a fixed-LCG phrase grid) — past the CBO's index crossover (~3,000 rows with fresh stats), so proof 1 captures the IVF vector index driving the plan - proof 1: DBMS_XPLAN 'BASIC +PREDICATE' exhibit + 4 new assertions proving the relational predicates evaluate on their scans below the top-10 stopkey (filter first, then rank); module now 23 assertions - init gathers optimizer stats explicitly (deterministic index-vs-scan choice) - arch-aware mongosh install (aarch64 podman builds); CONTAINER_ENGINE env in the validator; healthcheck window widened for the longer first-boot init - validated: cold builds green on aarch64 podman and x86_64 CI (43 assertions) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…filter-first proof, HVI) # Conflicts: # labs/converged-database-lab/docker/init/08-vector-model.sql
…ves to subject The documented merge-time collision (article 3 header MERGE NOTE): a SEARCH index and the HVI's text component are the same indextype and cannot share the body column (ORA-29879), and the HVI's text component does NOT sync on commit (verified empirically 2026-07-24 — a committed probe row was not CONTAINS-findable through the HVI). So on main: - 05-text-vector: ticket_text_idx (SYNC ON COMMIT) now indexes SUBJECT - 09-hybrid-index: drop of ticket_text_idx removed — HVI owns body, no collision - module 02 proof 3: marker token moves to the subject; CONTAINS(subject, ...) (the article/02 branch keeps the published body-marker form) - READMEs: index table + module map updated (module 03 on main, 23 assertions) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Hub main had removed the model layer along with 08-vector-model.sql; the 3-way merge honored that deletion while keeping the branch's mongosh arch fix, yielding an image with no /opt/oracle/models — LOAD_ONNX_MODEL failed with ORA-22288 at init. Restored the validated branch Dockerfile verbatim. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
rhoulihan
force-pushed
the
merge/article-03-lab-into-main
branch
from
July 24, 2026 15:29
d7461b3 to
cdd1a3a
Compare
oracle-contributor-agreement
Bot
added
OCA Verified
WSPluta
approved these changes
Jul 24, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Brings the article-3 companion module to main and reconciles the documented modules-02+03 collision.
What's in
DBMS_XPLAN +PREDICATEpredicate list proving filter-before-rank; vector read-after-write; VPD-governed retrieval; native Hybrid Vector Index hybrid search.CONTAINER_ENGINEsupport in the validator, healthcheck window widened for the longer first-boot init (~8–10 min).Reconciliation (the merge-time decision documented in the article-3 header)
A SEARCH index and the HVI's text component are the same indextype (ORA-29879, one per column), and the HVI's text component does not sync on commit (verified empirically 2026-07-24). On main:
ticket_text_idx(SYNC ON COMMIT) moves to subject for module 02's transactional read-after-write proof, the HVI owns body text search, and module 02's probe carries its marker in the subject. Thearticle/02branch keeps the published body-marker form its snippets byte-match.Validation
Cold podman build of this exact tree: init clean, validator ALL PASS — 12 scripts, 61 assertions across modules 01+02+03 (module 02's reconciled proof included). Module-03 content previously CI-green on x86_64 (working repo) and aarch64 podman.
🤖 Generated with Claude Code