What a lambda really is, how to pick the right interface from java.util.function by shape, why primitive versions exist, and the five compile errors every beginner meets. Every program was compiled and run on Java 25 and JDK 27.
RedisTemplate vs StringRedisTemplate, @RedisHash, Pub/Sub, TTL and Redis as the Spring cache on Spring Boot 4.1, with the JDK-serialisation default and the Jackson 3 fix shown in redis-cli.
Java 27 reached GA on 15 September 2026. Every one of its nine JEPs run on real JDKs, plus the ten JEPs of Java 26 for anyone jumping from the 25 LTS: G1 becoming the default in a one-CPU container (26 vs 27), 8-byte object headers measured with JOL, the post-quantum TLS handshake, JFR redaction, preview APIs that stopped compiling, removed JVM flags, and the Java 26 recap.
A naive @SchemaMapping resolver measured at 21 SQL statements for 20 books versus a @BatchMapping resolver's flat 2, via DataLoader batching and one .distinct() call. Plus a dangling foreign key nulling an entire GraphQL response through non-null propagation, byte-identical under both resolver strategies.
springdoc-openapi 3.1.1 silently merges same-path, different-version Framework 7 handlers into one oneOf operation with an arbitrary operationId. A working per-version fix with GroupedOpenApi and OpenApiCustomizer, and why the officially-versioning-supported functional-endpoint path documents fewer versions, not more.
Spring Data JDBC and Spring Data JPA, modelled against the identical Order/Customer domain and measured through the same SQL-statement counter, so every claim about N+1 queries, lazy loading, and the aggregate boundary is a number you can reproduce.
A real four-deploy expand-contract migration built with Spring Boot 4.1 and Flyway, proven live against two running replicas with a continuous-traffic load generator: 99.98% success, plus a genuine NOT NULL trap, a backfill-window bug, and H2 silently dropping a concurrently committed row during ALTER TABLE — confirmed to be a database-specific behavior, not a flaw in the technique.
A from-source comparison of Flyway 12.4.0 and Liquibase 5.0.3 on Spring Boot 4.1.1: checksum validation, out-of-order migrations, baselining an existing database, why Flyway Community has no working rollback, Liquibase's very different rollback and locking model, a real Liquibase 5.0.3 defect found and verified during testing, the FSL license change and its fallout at the ASF and Keycloak, and what happens when you run both tools against one database. Every claim traces to a companion project you can clone and run yourself.
A partitioned Spring Batch step scales a 10-million-row job across worker threads — until gridSize doesn't mean what you think, a rejected partition gets stuck forever, and the real speedup tops out well short of 2x.
Spring Boot 4.1 ships Spring Batch 6.0. This article builds a real chunk-oriented import job against a CSV of products, deliberately breaks it on a duplicate row mid-run, and shows exactly what a restart resumes (and what it doesn't) against real captured output: a chunk that commits as a whole or not at all, a second JVM that resumes at row 41 instead of row 1, skip-instead-of-fail as the alternative to restarting, and the resourceless job repository that silently forgets a restart ever happened.