jiangxt2 opened a new pull request, #13430: URL: https://github.com/apache/gravitino/pull/13430
### What changes were proposed in this pull request? Add the public `DATA_SKIPPING_VECTOR_SIMILARITY` index type and ClickHouse support for discovering, creating, altering, and reloading `vector_similarity` indexes with validated HNSW properties and normalized defaults. ### Why are the changes needed? Expose ClickHouse 25.8+ vector similarity indexes through Gravitino's relational index API while preserving index fields, supported parameters, and effective granularity on reload. Fix: #11935 ### Does this PR introduce _any_ user-facing change? Adds `Index.IndexType.DATA_SKIPPING_VECTOR_SIMILARITY` for ClickHouse 25.8 or later, with `L2Distance` and `cosineDistance`, HNSW dimensions, optional quantization and construction parameters, and a default granularity of 100000000. ### How was this patch tested? - `./gradlew rat` and Spotless checks for `common`, `integration-test-common`, and `catalog-jdbc-clickhouse` passed. - `./gradlew :common:test --tests 'org.apache.gravitino.json.TestJsonUtils' -PskipITs` passed (13 tests). - `./gradlew :catalogs-contrib:catalog-jdbc-clickhouse:test -PskipITs -PskipDockerTests=true` passed (135 tests). - `./gradlew :catalogs-contrib:catalog-jdbc-clickhouse:test -PskipITs -PskipDockerTests=false` passed (179 tests, including ClickHouse engine restart coverage). - With `GRAVITINO_CLICKHOUSE_TEST_IMAGE=clickhouse:25.8.14.17`, the full `CatalogClickHouseIT` passed (63 tests), including explicit HNSW zero parameters, and the full `CatalogClickHouseClusterIT` passed (21 tests). - `./gradlew :catalogs-contrib:catalog-jdbc-clickhouse:build -x test` and `./gradlew :docs:build` passed. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
