Confluence Connector - Deployment — Unique AI Documentation

Confluence Connector - Deployment

Content

The Confluence Connector is delivered as part of the Unique Connectors repository. Each release includes:

Container Image

Container images are available from GitHub Container Registry:

ghcr.io/unique-ag/connectors/services/confluence-connector:<version>

The image contains the application code plus all necessary runtime dependencies.

Helm Chart

The Helm chart wraps the backend-service subchart (~10.1.0, aliased as connector), so image, env, resources, and volumes are nested under the connector key. Tenant-specific configuration lives under connectorConfig.

Installation

Chart and app versions are maintained in:

Helm (via helm-git plugin):

Prerequisite: The helm-git plugin is required for git+https:// chart references. Install it with:

bash

helm plugin install https://github.com/aslafy-z/helm-git

helm repo add cfc git+https://github.com/Unique-AG/connectors@services/confluence-connector/deploy/helm-charts?ref=<release-tag>&depupdate=1
helm install confluence-connector cfc/confluence-connector \
  --version <version> \
  --namespace confluence-connector \
  --create-namespace \
  --values values.yaml

Helm Values Example

For the current Helm defaults, use values.yaml. For chart metadata, use Chart.yaml. For detailed configuration reference, see Configuration.

connector:
  image:
    repository: ghcr.io/unique-ag/connectors/services/confluence-connector
    tag: "<version>"
  env:
    LOG_LEVEL: info
  envVars: []
  resources:
    limits:
      memory: 1Gi
    requests:
      cpu: 1
      memory: 512Mi

# Tenant configuration (rendered into a ConfigMap and mounted as YAML)
connectorConfig:
  enabled: true
  tenants:
    - name: default
      confluence:
        instanceType: cloud
        cloudId: "your-cloud-id"
        baseUrl: https://your-domain.atlassian.net
        auth:
          mode: oauth_2lo
          clientId: "your-oauth-client-id"
          clientSecret: "os.environ/CONFLUENCE_CLIENT_SECRET"
        apiRateLimitPerMinute: 1200
        ingestSingleLabel: ai-ingest
        ingestAllLabel: ai-ingest-all
      unique:
        authMode: cluster_local
        ingestionServiceBaseUrl: "http://node-ingestion.<namespace>:8091"
        scopeManagementServiceBaseUrl: "http://node-scope-management.<namespace>:8094"
        apiRateLimitPerMinute: 100
        serviceExtraHeaders:
          x-company-id: "your-company-id"
          x-user-id: "your-service-user-id"
      processing:
        concurrency: 1
        scanIntervalCron: "*/15 * * * *"
      ingestion:
        ingestionMode: flat
        scopeId: "your-scope-id"
        storeInternally: enabled

Terraform Modules

Each release contains a Terraform module for secret provisioning:

Secrets Module

The secrets module provisions Azure Key Vault secret placeholders for the connector's credentials (OAuth client secret, or PAT for Data Center below 10.1 only). The secret values are set to <TO BE SET MANUALLY> on creation and must be populated by the operator. The lifecycle.ignore_changes block ensures Terraform does not overwrite manually set values on subsequent applies.

module "confluence_connector_secrets" {
  source = "github.com/Unique-AG/connectors//services/confluence-connector/deploy/terraform/azure/confluence-connector-secrets"

key_vault_id = var.key_vault_id

# Optional: override default secret placeholders
  # secrets_placeholders = {
  #   confluence-connector-pat-dc              = { create = true, expiration_date = "2099-12-31T23:59:59Z" }
  #   confluence-connector-client-secret-cloud = { create = true, expiration_date = "2099-12-31T23:59:59Z" }
  # }
}
Variable Type Required Description
key_vault_id string Yes The ID of the Azure Key Vault where secrets will be created
secrets_placeholders map(object) No Map of secret names to create. Each entry supports create (bool) and expiration_date (string). Defaults include confluence-connector-pat-dc and confluence-connector-client-secret-cloud

Output:secret_names - list of created Key Vault secret names (each prefixed with manual-).

Requirements: Terraform ~> 1.10, hashicorp/azurerm provider ~> 4.

Releases

All releases (including pre-releases) are available at: https://github.com/Unique-AG/connectors/releases

Version Numbering

The connector follows Semantic Versioning:

Pre-Release Versions

Pre-release versions are marked with suffixes like -alpha or -beta:

2.0.0-alpha.1
2.0.0-beta.1
2.0.0  (GA)

Release Tags

Release tags in the repository follow the format:

confluence-connector@<version>

Version Support and Maintenance Policy

The Confluence Connector follows Unique's standard release lifecycle and maintenance expectations. The canonical source is Upgrade and Release Process.

In short, plan operations around the latest release line and the previous supported line. Always verify current support boundaries and rollout expectations in the linked policy before upgrading production environments.

Compatibility

The Confluence Connector is compatible with:

Unique Platform Compatibility

Connector Version Minimum Unique Version
< 2.1.0 TBD
>= 2.1.0 node-ingestion - 2026.24.0+

Connector 2.1.0 introduces inlining of page images as base64 data: URIs (see Configuration -- Image Attachments). Extracting searchable text from those inline images is performed platform-side and only exists from Unique 2026.24.0. On earlier platform versions the connector still ingests the page, but embedded images contribute no searchable content. The platform-side extraction is additionally gated behind two company-scoped feature flags (FEATURE_FLAG_ENABLE_HTML_INLINE_IMAGE_EXTRACTION_UN_20936 and FEATURE_FLAG_ENABLE_MULTI_FILE_IMAGE_CONTENT_EXTRACTION_UN_20936), both of which must be enabled for the company. The connector must also be configured with the extraction model via ingestion.pageIngestionConfig (see Configuration -- Image Attachments).

Upgrading

Pre-Upgrade Checklist

  1. Review release notes for breaking changes
  2. Backup current Helm values
  3. Test upgrade in non-production environment
  4. Plan maintenance window if needed

Upgrade Process

# Review changes before upgrading
helm diff upgrade confluence-connector cfc/confluence-connector \
  --version <new-version> \
  --namespace confluence-connector \
  --values values.yaml

# Perform upgrade
helm upgrade confluence-connector cfc/confluence-connector \
  --version <new-version> \
  --namespace confluence-connector \
  --values values.yaml

Rollback

# List release history
helm history confluence-connector -n confluence-connector

# Rollback to previous version
helm rollback confluence-connector <revision> -n confluence-connector

Troubleshooting

Pod Not Starting

  1. Check pod events:
kubectl describe pod -l app=confluence-connector -n confluence-connector
  1. Check logs:
kubectl logs -l app=confluence-connector -n confluence-connector

Configuration Validation Errors

The connector validates all tenant configuration at startup. Check the pod logs for specific validation error messages. See Configuration for required fields and defaults.

Authentication Errors

Network Connectivity

See the network requirements table in the Operator Guide for the full list of required endpoints.

See FAQ for more troubleshooting guidance.