Skip to content

Site: bring the landing page, docs index, compare page and README up to v1.8.0 - #308

Merged
colinmollenhour merged 1 commit into
mainfrom
release/v1.8.0-site
Nov 18, 2026
Merged

colinmollenhour merged 1 commit into
mainfrom
release/v1.8.0-site

Conversation

@colinmollenhour

Copy link
Copy Markdown
Collaborator

AI · By: Claude Code with Claude Fable 5.2

Summary

The public site described the product as of early November. This brings the landing page, the docs index, the compare page and the README up to what v1.8.0 ships, and removes claims that stopped being true. It has to merge before the v1.8.0 tag, because bloodraven.dev is built from the release tag.

Why

The site deploys when a release is published. Features that shipped since (paced rollouts, the v1beta1 API, storage growth, the workload dashboard, standby clusters, the kubectl plugin) were in the reference docs but not on the pages a new visitor reads.

Changes

Landing page

  • Three new cards under Features: planned switchover, paced operator upgrades, and day-2 tooling (the kubectl plugin, six dashboards, the alert runbook map).
  • Two new cards under Data: storage that grows in place, and recovery into another cluster with MysqlStandbyCluster.
  • Two new points in the control-plane section: the versioned API with migrate-storage, and the operator's check at start.
  • "One CRD" is gone. There are five kinds. The headline now reads "One controller. One manifest. One loop.", and the stat reads "1 manifest per failover group".
  • Chaos scenario count: 60+ becomes 70+ (74 are registered).

Docs

  • Docs index: a table of the five kinds replaces "a single CRD"; six more standout features; 30+ scenarios becomes 70+.
  • Known limitations, API maturity: rewritten for v1beta1. It said fields may still change "before v1beta1" and that no migration contract exists.
  • Multi-site: a DisasterRecovery site is not a planned-failover target; the text said manual promotion "would use the forthcoming planned-failover API".
  • Multi-cluster DR: internal wishlist references and "available today on main" removed; what is shipped and what is planned is stated plainly.

Compare page (figures recounted from the code, as its header asks)

Figure Was Now
Registered chaos scenarios 62 74
Nightly scenarios (release profile) 28 29
Component tests / envtest tests 190 / 57 320 / 171
Grafana dashboards 5 6
Point-in-time recovery "refuses to replay across sites after a failover" replays in order across sites (fixed in v1.7.0)

README: the same six standout features and the scenario count.

Screenshots

None attached: the only image host available here is a public bucket, and the pages are public within the hour anyway. I rendered the landing page locally at 1280 px and inspected the new cards, the stats row and the control-plane points; the grids fill evenly (six cards in the three-column section, six in the two-column section).

Test Plan

  • npm test in site/: 38 passed.
  • Local nuxt dev: /, /compare, /docs, /docs/get-started/known-limitations and /docs/architecture/multi-cluster-dr return 200 with the new text.
  • Not run locally: the production build and the llms-full.txt check. CI's Docs / Build job runs both.

Not done

  • Other docs pages still cite internal wishlist numbers (operator availability, operations overview, upgrade policy, security model, durability). They are not about features, so I left them.
  • The compare page's "About 51 scenarios inject a failure" was not recounted.
  • "Sixty pages of docs" on the landing page: there are 54.

@colinmollenhour
colinmollenhour merged commit b418a37 into main Nov 18, 2026
28 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant