Source Baseline
| Reference | Revision or version |
|---|---|
| Previous release source tag | release-3.0.0, 3acb048b3f6e209d6d1e5ac54efcbd30f1b25c57 |
| Community trunk reviewed on 2026-09-07 | 94c6389a96b38bccef0b6a08269481a086b63ca1 |
| Monitoring target baseline, PR #4182 | 4e95d2e33493ac934d7d98a14a81d86c0f1bc0c4 |
| Multi-cluster follow-up reviewed on 2026-09-22 | fc07b5cb873dc154c48349085494e664baf6b9d1, including AMBARI-26656 / PR #4216 |
| Maven development version in the reviewed trunk | 3.1.0.0-SNAPSHOT |
For this documentation, PR #4182 is included in the intended 3.1.0 feature baseline. A release candidate must contain the selected community changes and monitoring implementation together; a reference SHA in this table is not an instruction to replace newer community work with an older feature branch.
The website release label is 3.1.0. The four-part Maven/package development version does not identify an intervening public 3.0.x release.
How Changes Were Selected
The review uses source ancestry, not only commit dates. A feature branch can contain older-dated commits that first become part of the mainline after 3.0.0. The reviewed trunk has 181 commits reachable after the previous release tag; the release notes group their user-visible impact rather than listing every test or merge separately.
Maintainers can reproduce the comparison against the pinned reviewed revision:
git log --first-parent --date=short --format='%h %ad %s' \
release-3.0.0..94c6389a96b38bccef0b6a08269481a086b63ca1
git diff --stat release-3.0.0..94c6389a96b38bccef0b6a08269481a086b63ca1
Primary Evidence
| Topic | Source |
|---|---|
| Monitoring control/data planes and replacement boundary | Telemetry architecture |
| VictoriaMetrics deployment topology and configuration | Managed Stack service |
| New monitoring database model | UpgradeCatalog310 |
| Managed Java dependencies | Project POM, build policy |
| Python runtime and distribution contract | Agent metadata, Agent packaging, effective implementation audit |
| React routing and permissions | Routes, primary UI selection |
| Selected monitoring runtime captures | Three-node evidence |
Evidence Boundaries
The monitoring evidence includes a three-node Rocky Linux 8 ARM64 deployment with VictoriaMetrics Server, VMAGENT, VMAUTH, and selected service/host dashboards. It does not certify every distributed topology or every failure mode.
The Python modernization record includes Linux x86_64 RPM and deployment validation, plus separate staged-artifact checks for other ABI/architecture profiles. Configuration support for another profile is not the same as a published and qualified binary for every distribution.
The React parity baseline distinguishes implementation, static comparison, and runtime validation. Its older review snapshot is not a current test certificate and must not be converted into an unconditional statement that every role, SSO mode, custom Stack, and recovery path has passed. The React guide lists the implemented workflow surfaces and acceptance boundaries.
Documentation build and browser checks validate this website's routes, translations, labels, and assets. They do not execute an Ambari cluster upgrade, Maven/RPM production build, or monitoring failure drill.
Multi-cluster Follow-up
The multi-cluster guides use the newer pinned trunk revision above for cluster isolation, durable wizard recovery, Admin management and managed HBase dependencies. This is a focused source follow-up, not a rerun of the original whole-release review or a fresh RPM acceptance run.
The guides distinguish final packaged independent-cluster evidence from the earlier managed-HBase deployment with overlays. They also document the remaining real-KDC, detach/data-retention, fault-injection and broker-revocation gates. The operations guide lists the evidence still needed before publishing final support claims.
Runtime Mpack And Console Follow-up
The runtime-store guides add a separately identified development follow-up reviewed on 2026-09-29:
| Input | Reviewed snapshot |
|---|---|
| Ambari implementation | Branch AMBARI-26663, commit 3a71190847503b033036833d1d49ba0ce325ca41 |
| Independent reference store | Commit c10a271; exact definition versions selected by release.json |
| Reference deployment scope | Rocky Linux 8/aarch64, systemd, basic non-HA services |
| Console acceptance scope | Desktop; grouped service selection, content editors, appearance/navigation and monitoring interactions |
The implementation snapshots were reviewed in development checkouts. They are not presented as newly published Apache source tags, downloadable ASF binary releases, or proof that every 3.1 build contains these changes. Obtain a build and store supplied together, and inspect the published manifest/API capabilities before using the workflow.
Evidence inputs include the core checkout's docs/mpack/http-api.md, docs/mpack/corrective-implementation.md and docs/frontend-refactor/react-current, plus the store's release.json, service READMEs, docs/content-configuration.md and docs/content-configuration-acceptance.md. The later content-configuration acceptance supersedes earlier definition-version examples and the old non-installable MinIO draft.
The snapshot has basic runtime checks for the ten selected services and targeted configuration failure/recovery checks. It does not establish general HA, multi-platform, all-role, data-migration, or production-load certification. Website build/browser tests establish documentation behavior only.
Start with the store overview, service matrix, and workspace guide.
Before Publishing A Release
Update this baseline after the release candidate is selected. Record final source tags, signed artifact/checksum locations, package target matrix, supported upgrade paths, test results, and the release vote outcome. Only then replace the preview designation with the released version.