Skip to main content

Connectors

The Connectors page is the catalog of every data source the platform can ingest from. Each card represents one connector type, shows whether the tenant has any configured instances, and exposes setup instructions or a dashboard link for that source.

Status Vocabulary​

Three states per card:

  • active — One or more configured instances are reporting data
  • idle — Configured instances exist but haven't reported data recently
  • not configured — No instances configured for this connector type

Categories​

Connectors are grouped into sections so the page is scannable:

  • Cloud providers — AWS, Azure, GCP, Mongo Atlas
  • Compute — Hosts, containers, Kubernetes
  • Databases — Postgres, MySQL, SQL Server, Mongo, Cosmos DB, Azure SQL, Oracle, and the rest of the database catalog
  • Middleware — Kafka, RabbitMQ, Redis and other message brokers / caches
  • Web servers — Nginx, Apache, IIS, Tomcat
  • Observability backends — OpenTelemetry, Splunk, Datadog
  • ITSM — PagerDuty, ServiceNow, Jira, Opsgenie

Each section can collapse independently — useful for tenants who don't use a whole category.

Setup Dialog​

Clicking the help icon on a card opens the setup dialog. It renders a markdown setup guide tailored to the connector — credentials needed, agent install steps, configuration snippets, and verification checks. For Kubernetes the dialog includes a seeder panel that materialises the hundred-rule Kubernetes alert catalog for each registered cluster in one click.

Each card has two link targets:

  • Open dashboard — Jumps to the page that surfaces this connector's data (e.g. the Kubernetes connector card → the Kubernetes page)
  • Configure — Jumps to the matching setup page or settings tab to add a new instance

Best Practices​

  • Walk the catalog the first time a tenant is set up — every card you make active is a data source the team can now alert and report on.
  • The setup dialog's markdown is the canonical reference for credentials and permissions; don't rely on tribal knowledge for production connector setup.
  • The K8s Catalog Seeder is idempotent — re-running it on a cluster that already has the rules reports 0 created, 100 skipped, so it's safe to click during an environment promotion.
  • Keep an eye on idle cards in production — they usually indicate a connector that's lost its credentials or its source has gone quiet.
  • Cloud — Cloud-provider connector instances and ingest pipelines
  • Databases — Database connector targets and discovered cloud databases
  • Alerts — Channels and rules that consume connector data
  • Settings — Tenant-level ingest pipelines (alert ingest, OTel ingest, incident ingest)