.png)
Narvar Integration Guide
Connect Narvar with commerce, ERP, fulfillment, carrier, and customer-service systems through REST APIs, selected callbacks, and orchestrated synchronization workflows.
Narvar integration options at a glance
Narvar integrations primarily use REST APIs to exchange orders, shipments, packages, tracking events, returns, and notification-related information. Selected Narvar products or tenants may also provide outbound callbacks or event notifications, although coverage, payloads, signing, and retry behavior must be confirmed. Enterprise implementations may use approved scheduled data feeds or managed file transfers, but exact formats and availability require tenant confirmation. Authentication is account-specific and may use an API key, bearer token, client credentials, or another provisioned method. Martini can consume Narvar APIs, expose REST endpoints, orchestrate workflows, transform payloads, and implement scheduled reconciliation with validation, checkpoints, retries, and idempotency controls.
Common Narvar integration patterns
Common Narvar data objects used in integrations
Authentication and security considerations
Tenant-specific authentication
Narvar credentials are provisioned during partner or tenant onboarding, but the exact public authentication scheme is not conclusively verified. Confirm whether the tenant uses an API key, bearer token, client credentials, or another account-specific method.
Secure credential handling
- Store Narvar credentials in Martini secrets rather than workflow definitions or mappings.
- Keep credentials and scopes environment-specific.
- Confirm token or key rotation, permitted resources, and expiration behavior.
- Separate authentication failures from validation, rate-limit, and server failures.
Callback protection
If Narvar callbacks are enabled, confirm signature or shared-secret requirements, validate authorization before processing, acknowledge requests according to the delivery contract, and avoid logging credentials or unnecessary customer information.
Operational considerations for Narvar integrations
Throughput and pagination
Confirm Narvar tenant limits, page sizes, pagination behavior, payload limits, and any retry headers. Use controlled concurrency, exponential backoff, and respect Retry-After when provided.
Idempotency and ordering
Use stable order, shipment, package, tracking, and return identifiers. Tracking events can arrive late, out of order, or more than once, so compare event timestamps and status precedence instead of assuming the latest received event is the latest business state.
Checkpoints and reconciliation
Polling workflows should use durable timestamps, cursors, or business checkpoints. Reconciliation provides recovery for missed callbacks, incomplete submissions, and unknown outcomes after network timeouts.
Schema and data quality
Validate carrier codes, tracking numbers, identifiers, addresses, quantities, timestamps, time zones, and return references. Use explicit mappings and monitor new or unknown fields as Narvar products and carrier integrations evolve.
Observability and testing
Record correlation identifiers, source IDs, HTTP status, error payloads, workflow execution IDs, retry counts, processing timestamps, and final synchronization state. Test representative shipment, delivery-exception, and return payloads in a non-production environment.
Why use Martini instead of scripts or point-to-point integrations?
Coordinate the complete integration
Scripts and point-to-point interfaces often leave authentication, transformation, retries, reconciliation, and operational visibility scattered across implementations. Martini brings those concerns into workflows and APIs that can be maintained as reusable integration assets.
Handle Narvar-specific variability
Narvar object schemas, callback coverage, authentication, and feed options may vary by product and tenant. Martini can isolate those differences behind explicit mappings, validation rules, environment configuration, and controlled workflows.
Support multiple operating modes
The same integration approach can combine API-led order and shipment submission, selected callback handling, scheduled polling, reconciliation, and approved file processing without requiring direct database access or an assumed connector.
Improve reliability and supportability
- Apply retry, throttling, idempotency, and checkpoint logic consistently.
- Separate transient, authentication, validation, and business failures.
- Preserve correlation identifiers and per-record outcomes for investigation.
- Expose controlled APIs for surrounding commerce, ERP, warehouse, and service systems.