Skip to main content

Core Concepts

  • Client: NexlaClient orchestrates authentication and request lifecycle; see nexla_sdk/client.py.
  • Credentials: CredentialsResource provides reusable access to external systems and is referenced by sources and destinations.
  • Resources: High-level managers for flows, sources, destinations, nexsets, lookups, users, organizations, teams, projects, notifications, and metrics (nexla_sdk/resources/*).
  • Models: Pydantic request/response schemas (nexla_sdk/models/*) keep payloads type-safe and serializable.
  • Errors: A rich hierarchy (nexla_sdk/exceptions.py) surfaces context such as operation, resource type, and retry hints.
  • Telemetry & Pagination: Optional OpenTelemetry spans wrap HTTP calls, and every resource inherits consistent pagination helpers.

Flow Topology​

flowchart LR
    Credentials["Credential\n(/data_credentials)"]
    Source["Source\n(/data_sources)"]
    Transform["Transform Step\n(Nexset.transform)"]
    Nexset["Nexset (Dataset)\n(/data_sets)"]
    Destination["Destination\n(/data_sinks)"]
    Monitor["Metrics & Runs\n(/metrics, flow runs)"]

    Credentials -->|data_credentials_id| Source
    Source -->|flow graph| Transform
    Transform --> Nexset
    Nexset --> Destination
    Source -.-> |detected data_sets| Nexset
    Nexset --> Monitor
    Destination --> Monitor
  • Credentials secure connection details and can be probed for structure/sample data (client.credentials.probe_*).
  • Sources ingest data using a connector source_type/connector_type and emit discovered nexsets; see nexla_sdk/models/sources/responses.py.
  • Transforms live on nexsets via NexsetCreate.transform and define schema-mapping logic before data is published downstream.
  • Destinations subscribe to nexsets and push data to sinks (sink_type).
  • Flows stitch nodes (source → nexset(s) → destination) and expose run metrics via client.flows.list(include_run_metrics=True) and client.metrics.*.