Skip to content

docs: surface Redpanda Connect pipelines for operator users (DOC-2275)#1808

Open
JakeSCahill wants to merge 8 commits into
betafrom
doc-2275-connect-pipelines-discovery
Open

docs: surface Redpanda Connect pipelines for operator users (DOC-2275)#1808
JakeSCahill wants to merge 8 commits into
betafrom
doc-2275-connect-pipelines-discovery

Conversation

@JakeSCahill

@JakeSCahill JakeSCahill commented Jul 17, 2026

Copy link
Copy Markdown
Contributor

DRAFT: hold until the Pipeline CRD ships in a released operator. Ships with operator v26.2.x.

Documents the Pipeline resource (redpanda-operator#1677) for operator users. Originally a discovery-only PR; now expanded into the full operator-side guide, per review feedback, covering every flow in the eng PR description with worked, e2e-validated examples.

Pages

  1. manage:kubernetes/k-connect-pipelines.adoc (nav: "Redpanda Connect Pipelines") — full guide:
    • When to choose the Pipeline resource over Kafka Connect
    • Enable the Connect controller (connectController.enabled + enterprise.licenseSecretRef)
    • First pipeline (inline configYaml), lint-gated startup, config/credential checksum rollouts, pause
    • Flow A — cluster-bound: cluster.clusterRef + userRef with a scoped User CR, the rendered-config injection contract (seed_brokers/tls/sasl + top-level shared client), user-keys-win escape hatch, ACL-scoping guidance
    • Flow B — external/BYOC: cluster.staticConfiguration with Secret/ConfigMap-backed TLS CA, inline SASL, CEL exclusivity with userRef
    • Flow C — per-pipeline cloud IAM: serviceAccountName + IRSA worked example (RDS IAM auth, no DB password anywhere)
    • Flow D — valueSources: all four source types (inline/secretKeyRef/configMapKeyRef/externalSecretRef) in one fan-out example
    • Workload knobs (image precedence, node pinning, extraInitContainers), monitoring (PodMonitor), license behavior, status/conditions reference, troubleshooting. Uses the standard shared:partial$enterprise-license.adoc notice and states that the license must include the Redpanda Connect product (matches the operator's IncludesProduct(ProductConnect) check).
  2. manage:kubernetes/k-autoscale-connect-pipelines.adoc (new, nav: nested "Autoscaling") — step-by-step autoscaling through the scale subresource, grounded in manifests validated e2e on EKS (HPA cpu 1→4; KEDA kafka-lag 1→6; KEDA prometheus input_received 1→3):
    • kubectl scale + raw /scale smoke test
    • HPA on cpu/memory (metrics-server, requests, behavior windows, real kubectl get hpa output)
    • KEDA: consumer-group lag trigger (with TriggerAuthentication for TLS/SASL listeners, partition-count capping), Prometheus trigger on Connect-emitted metrics (connectController.monitoring.enabled + PodMonitor), scale-to-zero
    • Interaction semantics: paused precedence, explicit replicas: 0, GitOps ownership of replicas, one autoscaler per pipeline
  3. Licensing overview (get-started:licensing/overview.adoc): new "Redpanda Operator enterprise features" section with its own table (matching the Redpanda/Console/Connect sections) covering both operator-gated features — Stretch Clusters (multicluster operator requires a valid license to start; LicenseValid condition + status.licenseStatus report expiry/violations on running clusters) and Connect pipelines (missing/expired license or missing Connect product entitlement → new pipelines report License: False and get no workload; running pipelines keep processing but stop receiving updates; recovery is automatic once a valid license is applied). Also adds a "Manage licenses for the Redpanda Operator" section (enterprise.licenseSecretRef chart values + per-cluster licensing via the Redpanda/StretchCluster enterprise field).
  4. What's new (get-started:release-notes/operator.adoc): "Redpanda Connect pipelines (beta)" and "Gateway API support for Redpanda Console" entries under the v26.2.x section.
  5. Console Gateway API HTTPRoute (deploy:console/kubernetes/deploy.adoc, from redpanda-operator#1329): new "Expose Redpanda Console with the Gateway API" section — Gateway API CRD/controller/Gateway prerequisites, Operator CRD + Helm gateway block examples, route verification (Accepted/ResolvedRefs, allowedRoutes gotcha), Ingress↔Gateway switching, and mutual-exclusivity note; both Ingress callouts route to it.
  6. CRD index row + TIP on Manage Kafka Connect: unchanged discovery surfaces, but the CRD index now routes to the local guide instead of the unpublished connect-docs page.

Build note

The two previously-expected broken-xref warnings for connect:install:kubernetes-operator.adoc are gone: all routing now targets the local pages. origin/beta has been merged into the branch.

Before un-drafting (GA batch)

  • Fill in the exact minimum operator version on k-connect-pipelines.adoc (TODO marker; expected v26.2.1)
  • When rp-connect-docs#457 publishes, add a Next-steps cross-link to connect:install:kubernetes-operator.adoc (comment markers in place)

Part of DOC-2275; eng feature is redpanda-operator#1677.

Preview pages

🤖 Generated with Claude Code

Three discovery surfaces for the 26.2 Pipeline CRD (DOC-2275):
- new Manage > Kubernetes page routing to the full guide in the
  Connect docs, positioned next to Manage Kafka Connect
- Pipeline row in the CRD reference index
- TIP on Manage Kafka Connect pointing at the first-class alternative
@netlify

netlify Bot commented Jul 17, 2026

Copy link
Copy Markdown

Deploy Preview for redpanda-docs-preview ready!

Name Link
🔨 Latest commit 44b1fe4
🔍 Latest deploy log https://app.netlify.com/projects/redpanda-docs-preview/deploys/6a63b618c749b00008c825a0
😎 Deploy Preview https://deploy-preview-1808--redpanda-docs-preview.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify project configuration.

@coderabbitai

coderabbitai Bot commented Jul 17, 2026

Copy link
Copy Markdown
Contributor

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 4856e5a6-54cc-488e-822a-580b379a98f4

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • 🔍 Trigger review
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch doc-2275-connect-pipelines-discovery

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands.

@JakeSCahill

Copy link
Copy Markdown
Contributor Author

Update: the Redpanda Connect operator support (Pipeline CRD, redpanda-operator#1677) is expected to slip past 26.2, most likely to 26.3. This PR stays in draft until the CRD ships in a released operator. Do not merge for the 26.2 GA on July 28.

@david-yu

Copy link
Copy Markdown
Contributor

PR is merged redpanda-data/redpanda-operator#1677, this will go out as Beta for the 26.2 date. Could you ensure that examples are provided with examples of how to use HPA and KEDA for autoscaling?

david-yu and others added 2 commits July 24, 2026 10:49
…censing, and what's new (DOC-2275)

Turn the discovery page into a full operator-side guide covering every flow
from redpanda-operator#1677: enable the Connect controller, first pipeline,
clusterRef+userRef cluster binding (with the rendered-config injection
contract), staticConfiguration for external clusters, per-pipeline cloud IAM
via serviceAccountName (IRSA worked example), and valueSources with all four
source types. Adds status/condition reference and troubleshooting.

Add a dedicated autoscaling page with step-by-step HPA (cpu/memory) and KEDA
(consumer-group lag, Connect metrics via Prometheus, scale-to-zero) examples,
all grounded in manifests validated end-to-end on EKS.

Document the no-license/expired-license behavior for pipelines in the
licensing overview, and add the operator what's-new entry (version section
marked TODO until the ship vehicle is confirmed).

Route the CRD index row to the local guide instead of the unpublished
connect-docs page, which removes the two expected broken-xref warnings.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@david-yu
david-yu marked this pull request as ready for review July 24, 2026 18:09
@david-yu
david-yu requested a review from a team as a code owner July 24, 2026 18:10
david-yu and others added 3 commits July 24, 2026 11:12
The feature ships with operator v26.2.x, not v26.3.x. Drop the provisional
v26.3.x section and its TODO.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…ing overview

Give the operator its own enterprise-features table, matching the Redpanda /
Console / Connect sections: Stretch Clusters (multicluster operator refuses to
start without a valid license; LicenseValid condition + status.licenseStatus
on running clusters) and Connect pipelines (moved out of the Connect table).
Add a 'Manage licenses for the Redpanda Operator' section covering
enterprise.licenseSecretRef and per-cluster licensing via the Redpanda and
StretchCluster resources.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…ect product entitlement

Replace the bespoke license sentence on the pipelines page with the shared
enterprise-license partial, and state everywhere the license is discussed
(notice, prerequisites, license behavior, troubleshooting, licensing overview
table) that the license must include the Redpanda Connect product — matching
the operator's IncludesProduct(ProductConnect) check.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add an 'Expose Redpanda Console with the Gateway API' section to the Console
Kubernetes deploy page (operator PR #1329): Gateway API CRD + controller +
Gateway prerequisites, Operator CRD and Helm tabs for the gateway block,
route verification (Accepted/ResolvedRefs, allowedRoutes gotcha), and
switching between Ingress and Gateway API. Both Ingress callouts now point
at the new section, and the operator v26.2.x what's-new gains an entry.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
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.

2 participants