How do I troubleshoot data errors in Lookout for Equipment?

Short description

To troubleshoot data ingestion job failures, run the DescribeDataIngestJob AWS Command Line Interface (AWS CLI) command.

Example:

aws lookoutequipment describe-data-ingestion-job yourJobID

The output provides details about the ingestion job, and includes the FailureReason field. The FailureReason field gives information on why the job failed.

Based on why the job failed, complete the steps in the following section.

Resolution

Note: If you receive errors when you run AWS CLI commands, then see Troubleshooting errors for the AWS CLI. Also, make sure that you're using the most recent AWS CLI version.

No valid data has been found across all the submitted files

When data is ingested, the schema doesn't allow for you to change the component name or sensor names in the files. If more data is ingested, then the data must be from the same schema. Otherwise, you get the preceding error message.

To resolve this error, review the schema of the input files and the project dataset. If the schemas are different, then you must create a new project with a matching schema.

Note: To check the project dataset schema, run the DescribeDataset AWS CLI command.

No file matches the provided input Amazon S3 bucket

If you receive the preceding error, then Lookout for Equipment doesn't have access to time series data from your equipment's sensors.

To supply this data to Lookout for Equipment, use on of the following methods to upload the data from Amazon Simple Storage Service (Amazon S3).

Store all sensor data in a single .csv file

It's a best practice to store all your sensor data into a single .csv file. The first column in the .csv file must contain the date and time of the measurements. All additional columns can contain data from specific sensors in each separate column.

Store all sensor data in multiple .csv files

You can use a separate .csv file for each sensor to store all your sensor data. The first column in the .csv file must contain the date and time of the measurements. The second column can contain data from the sensor.

Note: The file must be a .csv file, or else you get the following error: "Cannot create schema. Reasons: The first column must contain date and time data in a supported timestamp format."

An error occurred when calling the StartDataIngestionJob operation

If the AWS Identity and Access Management (IAM) role used for the ingestion job doesn't have the appropriate permissions, then you might see the following error:

"An error occurred (ValidationException) when calling the StartDataIngestionJob operation: Role: arn:aws:iam::AcountId:role/LookoutEquipmentCrossAccountS3Accesss has no Read Access to Input S3 bucket, or the bucket is not owned by the requester. bucket"

If you use a custom role, then update the IAM policy and trust relationship policy.

Example IAM policy:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:GetObject",
        "s3:ListBucket"
      ],
      "Resource": [
        "arn:aws:s3:::YOUR_BUCKET_NAME",
        "arn:aws:s3:::YOUR_BUCKET_NAME/*"
      ]
    }
  ]
}

Example Trust Relationship policy:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "Service": "lookout-equipment.amazonaws.com"
      },
      "Action": "sts:AssumeRole"
    }
  ]
}

Or, you can receive the preceding error if the Amazon S3 bucket is present in another account. To resolve this issue, import your Lookout for Equipment datasets and models from a source AWS account to a target AWS account.