nizhikov opened a new pull request, #13685: URL: https://github.com/apache/ignite/pull/13685
Documentation update for the switch of the default SQL engine from H2 to Calcite. Stacked on #13670 (packaging swap); the first commits of this PR are the ones of #13670, only the last commit `eb450e9f298` is new. JIRA: https://issues.apache.org/jira/browse/IGNITE-29117 ### What changed * `SQL/sql-calcite.adoc` becomes the "SQL Engines" page: Calcite is the default and needs no setup; routing documents the engine names and the errors for an engine that is not available; a new "Legacy H2 Engine" section explains how to enable H2, how to keep the previous behaviour after an upgrade, lists the features implemented by H2 only, and maps every new error text to its cause and remedy. * `ignite-indexing` is no longer presented as the SQL prerequisite (`setup.adoc`, `sql-introduction`, `sql-api`, `indexes`, JDBC client driver, .NET deployment, snippets POM); `ignite-calcite` takes its place and `ignite-indexing` is described as the optional legacy engine. * JDBC `queryEngine` parameter and ODBC `QUERY_ENGINE` attribute are documented in the driver parameter tables. `distributedJoins`, `collocated`, `replicatedOnly`, `skipReducerOnUpdate` are marked as H2-only. * `COPY`, `SET STREAMING`, `TextQuery`, deprecated `SqlQuery` and the JDBC streaming section carry H2-only notes with the Calcite-era alternatives. * `distributed-joins.adoc`: Calcite executes non-colocated joins without a flag; the `setDistributedJoins` warning and the join limitations are scoped to H2. * SQL tuning guide: `EXPLAIN PLAN FOR` and the Calcite operator tree, `FORCE_INDEX`/`NO_INDEX` instead of `USE INDEX`, H2-only sections marked (`OR` selectivity, column count, `collocated`, query parallelism, skip reducer, on-heap row cache). The unlinked duplicate `SQL/sql-tuning.adoc` is removed and its three inbound links repointed. * "Only with Calcite" notes (hints, window functions, transactions, custom functions, session context, plan history) now read "not available with the legacy H2 engine". * TOC: "SQL Engines" placed right after the SQL introduction. ### Not covered here * The SQL reference pages (functions, data types, DML, conformance) still describe the H2 dialect. That rewrite is proposed as a separate ticket. * `installing-using-docker.adoc` keeps `ignite-indexing` in the default `OPTION_LIBS` until the Docker image repository is updated. * `spring-data.adoc` keeps `ignite-indexing`: the Spring Data extension may still issue `SqlQuery`. ### Verification The site was built locally with Jekyll (`ruby:3.3` container). All edited pages render, the new anchors resolve, and the build warnings are the pre-existing ones. 🤖 Generated with [Claude Code](https://claude.com/claude-code) -- 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]
