Connectors are software services installed on servers within your organisation's environment. They retrieve data from external systems on a schedule and transfer it securely to the platform for processing. Once a connector delivers a file, the platform's File Monitor detects it and triggers the appropriate import process automatically.
Three connector types are available. The Historian Connector retrieves time-series data from supported data historian systems. The Database (Measurement) Connector extracts data from relational databases via SQL queries. The File Connector monitors local or network directories for files and uploads them to the platform.
Each connector runs as a Windows service and includes a web-based administration interface for configuring queries, scheduling jobs, and monitoring activity.
How data transfer works
When a connector job runs, it retrieves data from the source system and generates a file containing the results. The file is uploaded to a secure storage location associated with the target platform instance. The File Monitor running within the platform instance then picks up the file and creates a process execution entry, passing the file as input to the configured import process. Once processing is complete, the file is removed.
If a file upload fails, the connector job is marked as failed. Retry behaviour is configurable via the connector configuration properties file.
Network requirements
The server hosting the connector must be able to reach specific platform service endpoints over the standard secure web port (443). These endpoints are used for authentication, file upload, and configuration retrieval. Access must be available for both production and test environments. Contact support for the specific endpoint addresses for your instance.
Deployment guidance
One installation per source and instance. Each connector installation sends data to a single platform instance and connects to a single source system. If you need to feed data from multiple sources, or to multiple instances, separate installations are required.
Redundant deployment. For scenarios where data retrieval must run several times a day and reliability is critical, install the same connector type on more than one server, each configured to run at slightly different times. If one connector fails to deliver data, the other can still complete the job. When redundant connectors are in use, the receiving platform instance must be configured to handle duplicate records correctly.
Certificates. Each connector authenticates with the platform using an X.509 client certificate. Certificates must be issued by a Certificate Authority (internal or external) and must support both Client Authentication and Server Authentication. The certificate subject should be 100 characters or fewer, must not contain wildcards, and should use the server hostname as the Common Name. A single certificate can be shared across multiple connector installations on the same server. The certificate's distinguished name appears as a user identity in the platform, so it must be assigned the appropriate roles and resource permissions. See Client Certificates for Connectors for the generation procedure.
Security scanning. Uploaded files are scanned before processing. Files that fail security checks are not processed.