### What Goes Wrong```
Expected NAD spec.config:
{
  "resourceName": "openshift.io/cx7anl244",  ✅ CNI needs this
  "pciAddress": "0000:02:01.2",              ✅ CNI needs this
  "type": "sriov"
}Actual NAD spec.config:
{
  "type": "sriov"
  # ❌ resourceName MISSING!
  # ❌ pciAddress MISSING!
}BUT: resourceName IS in metadata.annotations ✅
```

### Impact- ❌ Pod attachment fails
- ❌ All SR-IOV networking broken
- ❌ Tests timeout waiting for pod readiness
- ✅ Only manifests when creating NEW networks or after operator restart---

## When Bug Manifests

**NOT in normal operation** (pre-configured networks work fine)

**YES in these situations:**
1. Creating a NEW SriovNetwork resource
2. After operator restart/reinstallation
3. Comprehensive testing (like your tests!)
4. When NADs are regenerated

**Why**: Operator only creates NAD when you create SriovNetwork resource

---

## Root Cause

**Location**: `bindata/manifests/cni-config/sriov/` (template files)

**Issue**: Template placement logic puts `resourceName` in annotations but NOT in spec.config JSON

**Go Code**: ✅ CORRECT - properly prepares `data.Data["CniResourceName"]`

**Templates**: ❌ BUGGY - uses it in annotations but not in CNI config 

### Auto-Reproduction ToolFile: `reproduce_incomplete_nad_bug.sh`

**What it does:**
1. Creates test namespace
2. Creates SriovNetwork resource
3. Captures operator logs
4. Attempts to create test pod
5. Collects all evidence
6. Shows complete NAD output

**How to run:**
```bash
bash reproduce_incomplete_nad_bug.sh# Output: Complete NAD config in /tmp/
```     

 ### Scenario 1: Pre-configured Networks
```
Production Setup:
  Networks created long ago or pre-provided
    ↓
  Operator just uses them
    ↓
  ✅ No new NAD creation needed
    ↓
  ✅ Bug doesn't manifest
```
### Scenario 2: New Network Creation (Your Tests)
```
Your Test Setup:
  Create FRESH SriovNetwork resource
    ↓
  ❌ Operator generates NEW NAD with incomplete config
    ↓
  Try to attach pods
    ↓
  ❌ CNI plugin fails
```
### Scenario 3: Operator Restart
```
Production Scenario:
  Operator running (NADs exist)
    ↓
  Operator crashes/restarts
    ↓
  NADs regenerated
    ↓
  ❌ Regenerated NAD has incomplete config
    ↓
  ❌ Pods fail to attach
    ↓
  This is when bug appears in production

    pods attach ok