-
Notifications
You must be signed in to change notification settings - Fork 0
Description
Checked for duplicates
Yes - I've already checked
🐛 Describe the bug
When using the generated inventory CSV as a manifest for S3 Batch Operations with the pds_s3_data_tagger Lambda function, the batch operation fails with the following error:
Failed to parse task from manifest at byte offset 0. ErrorMessage: Unexpected number of task fields. Expected: 3. Observed: 7
The inventory CSV created from the "Determine unique data from existing inventory" task contains 7 columns, but S3 Batch Operations requires a CSV manifest with only 2-3 columns in a specific format: bucket name, object key, and optionally version ID.
🕵️ Expected behavior
The inventory CSV should be generated in a format compatible with S3 Batch Operations, containing only the required 2-3 columns:
- Bucket name
- Object key
- Version ID (optional)
This format is required for S3 Batch Operations to properly invoke the Lambda function for each object.
📜 To Reproduce
- Complete the "Determine unique data from existing inventory" task
- Use the generated CSV file as the manifest for an S3 Batch Operation
- Configure the batch operation to invoke the pds_s3_data_tagger Lambda function
- Attempt to run the batch operation
- Observe the error: "Failed to parse task from manifest at byte offset 0. ErrorMessage: Unexpected number of task fields. Expected: 3. Observed: 7"
🖥 Environment Info
- AWS S3 Batch Operations
- pds_s3_data_tagger Lambda function
- Generated inventory CSV (7 columns - incompatible format)
📚 Version of Software Used
pds_s3_data_tagger Lambda function (current version in main branch)
🩺 Test Data / Additional context
The generated inventory CSV needs to be transformed to match the S3 Batch Operations manifest format. According to AWS documentation, the CSV format must have either 2 or 3 columns in the following order:
- Column 1: Bucket name
- Column 2: Object key
- Column 3: Version ID (optional)
Reference: https://docs.aws.amazon.com/AmazonS3/latest/userguide/batch-ops-create-job.html
🦄 Related requirements
#196 (primary issue)
#187
#194
#197
For Internal Dev Team To Complete
⚙️ Engineering Details
To be filled by engineering team
🎉 Integration & Test
To be filled by engineering team
Metadata
Metadata
Assignees
Labels
Type
Projects
Status