If a pipeline fails whilst running, a FailedPipelineContext will be generated. This contains details of the failure. If the pipeline has a FailureEndpoint set, a call will be made to this endpoint with details of the failure.
FailedPipelineContext
A FailedPipelineContext can be retried, for example if you have since fixed a failing module, or an unavailable external service is now available.
If a pipeline was originally triggered by a DirectPost trigger, rather than data being added to a data collection, the FailedPipelineContext will not have stored the relevant data so cannot be retried.
See the Api Docs for more information
FailureEndpoint Notification
If a pipeline has a FailureEndpoint set, AireInsights will send a POST request to this url. The contents will be structured as follows:
{
"ErrorMessage": "Pipeline failed. Fail Reason: There was an error. Owner=testUser Pipeline=APipeline TransactionId=3a41f5a0-d841-4cd3-9565-6404d3df8728. A FailedPipelineContext with Id b8cdf7ac-9df3-4104-a403-4e9e36c07d9d has been saved.",
"FailReason": "There was an error",
"Owner": "testUser",
"PipelineName": "APipeline",
"TransactionId": "3a41f5a0-d841-4cd3-9565-6404d3df8728"
}
The request will be authorized by the API key of the Pipeline Owner
Those using AireInsights for high volumes of data processing should exercise caution and consider the effects of a high volume of notifications in the case of e.g. downtime from an outside service.
See the Pipeline Docs for more information