Appflow Scheduled Trigger Properties: first_execution_from not working on appflow

Terraform Providers AWS
1

I’m trying to set this property in my appflow in order to limit the initial pull, the standard in aws is to pull the last 30 days worth of data but I am trying to limit that.

Here is a simplified version of my terraform code:

Scheduled trigger to run the flow hourly

trigger_config {
trigger_type = “Scheduled”
trigger_properties{
scheduled {
schedule_expression = “rate(1hours)”
data_pull_mode = “Incremental”
first_execution_from = timeadd(timestamp(),“-24h”) #Pull starting from 24 hrs ago
schedule_start_time = timeadd(timestamp(),“4m”) # Start time is 4 mins from deploy
timezone = “America/Toronto”
}
}
}

I have tried with hardcoded values such as :
first_execution_from = “2025-03-03T18:27:00+00:00”

also tried using the dateformat function to ensure the correct format, but nothing is working.

when ever i run aws describe-flow on the flow it doesn’t show the first_execution_from field as a property on the flow.

can anyone provide me with a working example?

2

The first_execution_from argument requires an RFC3339 formatted timestamp. For example, here is the RFC3339 constant from the time package in the Go standard library.

"2006-01-02T15:04:05Z07:00"

If you have been able to pass the correctly formatted string and apply successfully, can you share more of the output from the execution (debug logs are helpful), and the output from the describe-flow CLI command where you’re observing the value is missing?

3

Hello Jar,

first of all i really appreciate you taking the time to answer me.

I tried using the value you suggested ( hardcoding just to see if it will work or not), and i got this error during terraform Apply:

│ Error: expected “trigger_config.0.trigger_properties.0.scheduled.0.first_execution_from” to be a valid RFC3339 date, got “2025-03-05T12:00:00Z05:00”: parsing time “2025-03-05T12:00:00Z05:00”: extra text: “05:00”

So i tried again using first_execution_from = “2025-03-05T12:00:00Z” and i was able to tf plan and tf apply successfully but when i ran the appflow it still pulled from the last 30 days.

Here is the output in the logs from the TF apply:
+ trigger_config {
+ trigger_type = “Scheduled”

      + trigger_properties {
          + scheduled {
              + data_pull_mode       = "Incremental"
              + first_execution_from = "2025-03-05T12:00:00Z"
              + schedule_expression  = "rate(1hours)"
              + schedule_start_time  = (known after apply)
              + timezone             = "America/Toronto"
            }
        }
    }

but when i run the describe flow this is what i see

"triggerConfig": {
    "triggerType": "Scheduled",
    "triggerProperties": {
        "Scheduled": {
            "scheduleExpression": "rate(1hours)",
            "dataPullMode": "Incremental",
            "scheduleStartTime": "2025-03-06T17:22:00+00:00",
            "timezone": "America/Toronto",
            "scheduleOffset": 0
        }
    }
},

i tried to create some debug logs but i didnt find they had anything helpful, i was unable to upload them as they are a .log file which is not supported on here. is there anything else i can provide?

4

At this point it may make sense to open an issue against the provider repo.

You should be able to attach debug logs as a .txt file, or link to a gist of the run output. Really the only lines we’re looking for are the logged CreateFlow API call to verify the FirstExecutionFrom field is populated in the request body. If you can extract just lines containing that API name and include them here that would also work.

If the provider is not sending the value correctly during the Create API call that would be a bug to fix on our end. If it is sending the value, the next step is likely a support case to AWS to determine why the FirstExecutionFrom value sent during the create operation is not returned in the corresponding Describe API call.