Oliver/ferrero-opentext
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
ferrero-opentext/Python-Version/MARKDOWN_DOCS/QUESTIONS_FOR_DAM_IT_TEAM.md

9.1 KiB
Raw Permalink Blame History

Questions for OpenText DAM IT Team - Metadata Inheritance

Date: October 29, 2025 Context: Content Scaling Workflow - Uploading derivative assets with metadata from master assets Issue: Need to copy metadata from master asset to derivative asset and update specific fields

Current Situation

What Works ✅

  • Simple Upload: Uploading new assets with 5 basic metadata fields works perfectly
    • FERRERO.FIELD.MKTG.ASSET TYPE
    • FERRERO.FIELD.FISCAL YEAR
    • MAIN_LANGUAGES
    • ARTESIA.FIELD.ASSET NAME
    • FERRERO.FIELD.STATE
  • Result: HTTP 202 Accepted, asset appears in folder, metadata applied correctly

What Doesn't Work ❌

  • Full Metadata Upload: When we send the complete metadata structure from a master asset (~128KB JSON)
    • Result: HTTP 202 Accepted, but asset does NOT appear in the folder
    • Behavior: Silent failure - no error message, just doesn't create the asset
    • Metadata Size: 128,912 bytes vs 2,000 bytes for simple structure

Our Goal

We want to implement a "derivative asset" workflow:

  1. Master Asset exists in DAM with complete metadata (16+ fields)
  2. Download master asset and all its metadata
  3. Store complete metadata in our database
  4. Agency creates derivative/localized versions
  5. Upload derivative with:
    • All metadata from master (inherited)
    • Updated fields from filename (language, country, asset name)
    • Result: Derivative has same metadata as master except for the fields we changed

Example Flow

Master Asset: 06_RAFFAELLO_MAESTRO_SD.mp4
  - Language: EN
  - Asset Type: TVC
  - Brand: RAFFAELLO
  - Agency: XYZ Agency
  - Fiscal Year: 2025/2026
  - [12+ other fields]

Derivative File: RAF_DE_de_XMAS_TVC_001_6S_16x9.mp4
  - Language: DE ← Updated from filename
  - Asset Type: TVC (same)
  - Asset Name: RAF_DE_de_XMAS_TVC_001_6S_16x9 ← Updated
  - Brand: RAFFAELLO ← Inherited from master
  - Agency: XYZ Agency ← Inherited from master
  - Fiscal Year: 2025/2026 ← Inherited from master
  - [All other fields inherited from master]

Questions for DAM IT Team

Question 1: Metadata Payload Size Limits

Q: Is there a maximum size limit for metadata payloads when uploading assets via POST /v6/assets?

Context:

  • Simple upload (5 fields): ~2KB payload → Works ✅
  • Full metadata (16+ fields): ~128KB payload → Silent failure ❌
  • Both return HTTP 202, but large payload doesn't create asset

What we need to know:

  • Is 128KB too large?
  • Is there a documented limit?
  • Is there a timeout for large payloads?

Question 2: Metadata Inheritance from Master Assets

Q: What's the recommended approach to copy/inherit metadata from one asset to another?

Options we're considering:

  1. Copy full metadata structure - Download master metadata and send it when uploading derivative
  2. Use asset reference - Reference the master asset ID and tell DAM to copy metadata
  3. Use template/copy API - Is there a dedicated endpoint for copying metadata between assets?
  4. Selective field upload - Only send the fields we want to change, DAM preserves others?

What we need:

  • Best practice for metadata inheritance
  • Any special API endpoints for this?
  • Example request structure

Question 3: Nested Category Structure

Q: When uploading metadata, do we need to preserve the categorized structure from the master asset?

Master metadata structure we're seeing:

{
  "metadata": {
    "metadata_element_list": [
      {
        "category_id": "1",
        "metadata_element_list": [
          {"id": "FERRERO.FIELD.MKTG.ASSET TYPE", "value": "..."},
          {"id": "MAIN_LANGUAGES", "value": "..."},
          ...
        ]
      },
      {
        "category_id": "2",
        "metadata_element_list": [...]
      }
    ]
  }
}

Questions:

  • Must we preserve the category structure?
  • Can we flatten to a single array of fields?
  • Does the API accept both formats?

Question 4: Updating Specific Fields

Q: Can we upload an asset with partial metadata and update specific fields later?

Scenario:

  1. Upload asset with basic metadata (5 fields) ← Works
  2. Then call an UPDATE endpoint to add/modify fields ← Does this exist?

Questions:

  • Is there a PATCH/UPDATE metadata endpoint?
  • Can we update metadata after asset creation?
  • What's the endpoint: PATCH /v6/assets/{id}/metadata?

Question 5: Field Type Requirements

Q: Are there specific field types that are required vs optional for the ECOMMERCE metadata model?

What we need:

  • List of required fields for ECOMMERCE model
  • Can we omit optional fields vs must send them with empty/default values?
  • Does omitting fields cause validation failures?

Question 6: Domain Values

Q: Some fields have complex domain value structures with field_value vs simple value. What determines which structure to use?

Example structures we're seeing:

// Structure A (domain value with field_value)
{
  "id": "FERRERO.MARKETING.FIELD.AGENCY NAME",
  "value": {
    "domain_value": true,
    "value": {
      "field_value": {"value": "0000000023"},
      "type": "com.artesia.metadata.DomainValue"
    }
  }
}

// Structure B (simple domain value)
{
  "id": "FERRERO.FIELD.MKTG.ASSET TYPE",
  "value": {
    "domain_value": true,
    "value": {"type": "string", "value": "heroimage"}
  }
}

Questions:

  • Which structure should we use for each field?
  • Is this defined in the lookup domains?
  • Can we query this via API?

Question 7: Tabular Fields

Q: How do we properly handle tabular fields like MAIN_LANGUAGES?

Current structure:

{
  "id": "MAIN_LANGUAGES",
  "parent_table_id": "FERRERO.TABULAR.FIELD.MAIN LANGUAGES",
  "type": "com.artesia.metadata.MetadataTableField",
  "values": [{"cascading_domain_value": false, ...}]
}

Questions:

  • Can tabular fields have multiple values?
  • Do we always use values array (plural)?
  • Is parent_table_id required?

Question 8: Silent Failures

Q: Why does the API return HTTP 202 (Accepted) but not create the asset when metadata is large?

Observations:

  • Small metadata (~2KB): HTTP 202 → Asset created ✅
  • Large metadata (~128KB): HTTP 202 → Asset NOT created ❌
  • No error message in either case
  • Both use identical upload structure, just different metadata content

Questions:

  • Is there async validation that can fail after 202 response?
  • How can we check if validation failed?
  • Is there a job status endpoint to check async upload status?
  • Should we poll for the asset to verify it was created?

What We've Tried

Attempt 1: Full Metadata from Master ❌

// Downloaded full master asset metadata
// Sent entire metadata_element_list structure (15 categories, 50+ fields)
// Size: 128KB
// Result: HTTP 202 but asset not created

Attempt 2: Simple 5 Fields ✅

// Created minimal metadata with 5 fields
// Size: 2KB
// Result: HTTP 202 and asset appears in folder

Technical Details

API Endpoint Used

POST https://ppr.dam.ferrero.com/otmmapi/v6/assets

Request Structure

POST /v6/assets
Content-Type: multipart/form-data

Parameters:
- asset_representation: {JSON structure}
- parent_folder_id: {folder_id}
- manifest: {upload_manifest JSON}
- files: {binary file data}

Authentication

OAuth 2.0 Bearer Token (client_credentials)
Working correctly (uploads succeed with simple metadata)

Target Folder

Final Assets Folder ID: ea0dbf86e13e3634895746d3a986558ec2eb8be1
Permissions: Write access confirmed

Recommended Approach (Need IT Team Guidance)

Option 1: Copy from Master Asset (Preferred)

Is there an API to say "create asset X with metadata copied from asset Y"?
Then we could update just the fields we need to change.

Option 2: Two-Step Process

Step 1: Upload with minimal metadata (5 fields) - Works
Step 2: PATCH/UPDATE to add remaining fields from master - Does this exist?

Option 3: Selective Metadata

Send only the fields that changed from master + references to master
Does DAM support referencing master metadata?

Option 4: Understand Payload Limits

If there's a size limit, we could:
- Send only essential fields
- Use metadata templates
- Reference external metadata

Sample Data Available

We can provide:

  • ✅ Working simple metadata JSON (2KB, 5 fields)
  • ✅ Full master metadata JSON (128KB, 50+ fields)
  • ✅ Successful upload request (works)
  • ✅ Failed upload request (202 but no asset created)
  • ✅ HTTP request/response logs

Priority Questions (Most Important)

  1. What's the proper way to inherit metadata from a master asset to a derivative?
  2. Is there a size limit on metadata payloads?
  3. Can we update metadata after asset creation? (PATCH endpoint)
  4. Why does HTTP 202 not always create the asset?

Contact Information

Project: Ferrero Content Scaling Workflow Developer: Dave Porter Environment: Production (ppr.dam.ferrero.com) Client ID: otds-OLV

Please advise on the best approach for metadata inheritance in derivative asset workflows. 🙏