An Azure analytics service that brings together data integration, enterprise data warehousing, and big data analytics. Previously known as Azure SQL Data Warehouse.
The available Synapse Apache Spark diagnostic emitters support only three destination types: Azure Log Analytics, Azure Storage, and Azure Event Hubs.
From the documented configuration options:
- Azure Log Analytics: uses a workspace ID plus a secret (Log Analytics key) or a secret stored in Azure Key Vault.
- Azure Storage: uses either an Access Key or SAS token, provided directly via
secretor indirectly via Azure Key Vault. - Azure Event Hubs: uses a connection string provided directly via
secretor indirectly via Azure Key Vault; there is also support for certificate-based service principal authentication for Event Hubs.
No documented option exists to:
- Send Synapse Spark logs directly to Geneva from Synapse diagnostic emitters without going through one of these three destinations.
- Use managed identity alone (without access key/SAS/Key Vault secret) to write Spark application logs (Log4j logs) from Synapse diagnostic emitters to Azure Storage.
For SFI-compliant routing, Event Hubs with secret or certificate-based authentication is the only documented path among the three emitter types. Any further cost optimization or Geneva integration would need to be implemented downstream of one of these supported destinations (for example, after Event Hubs), as there is no additional built-in, lower-cost, SFI-compliant Geneva target in the diagnostic emitter framework.
Similarly, for Storage, the documented authentication modes are AccessKey or SAS, either inline or via Key Vault; managed identity is not listed as a supported auth mechanism for the AzureStorage diagnostic emitter.
References: