Once migration plans are created, you often need to modify their configuration as migration requirements evolve. This chapter covers comprehensive plan modification techniques using kubectl-mtv patching capabilities, enabling dynamic updates without recreating plans.

Overview: Dynamic Plan Modification

Why Plan Patching?

Migration plans require updates throughout their lifecycle:

  • Environment Changes: Updated network configurations, storage requirements, or target namespaces
  • Migration Strategy Evolution: Switching between cold, warm, and live migration types
  • VM-Specific Customization: Individual VM requirements that emerge during planning
  • Hook Integration: Adding custom automation after initial plan creation
  • Performance Optimization: Adjusting convertor pod scheduling for better performance

Patching Capabilities

kubectl-mtv provides two primary patching mechanisms:

  1. Plan-Level Patching (patch plan): Modify plan-wide settings affecting all VMs
  2. VM-Level Patching (patch planvm): Customize individual VMs within the plan

Both approaches support comprehensive configuration updates verified from the command implementation.

Patching vs. Recreation

When to Patch:

  • Plan exists and contains complex VM selections
  • Need to preserve existing plan metadata and references
  • Making incremental configuration changes
  • Plan is referenced by other resources or automation

When to Recreate:

  • Fundamental changes to source/target providers
  • Major VM list modifications
  • Complete migration strategy overhaul

How-To: Patching Plan Settings

Basic Plan Configuration Updates

Migration Type Changes

Change migration strategies dynamically based on requirements:

1
2
3
4
5
6
7
8
9
10
11

# Switch from cold to warm migration
kubectl mtv patch plan --plan-name production-migration \
  --migration-type warm

# Enable live migration (KubeVirt sources only)
kubectl mtv patch plan --plan-name k8s-to-k8s-migration \
  --migration-type live

# Switch to cold migration for maximum reliability
kubectl mtv patch plan --plan-name critical-systems \
  --migration-type cold

Network and Storage Configuration

Update network settings and target configurations:

1
2
3
4
5
6
7
8
9
10
11

# Change transfer network for better performance
kubectl mtv patch plan --plan-name large-vm-migration \
  --transfer-network migration-network/high-bandwidth-net

# Update target namespace
kubectl mtv patch plan --plan-name dev-environment \
  --target-namespace development-new

# Modify description for better documentation
kubectl mtv patch plan --plan-name quarterly-migration \
  --description "Q4 2024 production workload migration to OpenShift 4.16"

Advanced Plan Configuration

Target VM Placement Updates

Modify where target VMs will be scheduled using various placement strategies:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# Add labels to all target VMs
kubectl mtv patch plan --plan-name production-apps \
  --target-labels "environment=production,migration-batch=2024-q4"

# Update node selector for hardware requirements
kubectl mtv patch plan --plan-name gpu-workloads \
  --target-node-selector "accelerator=nvidia-tesla-v100,node-type=compute"

# Apply advanced affinity rules using KARL (see Chapter 28 for syntax reference)
kubectl mtv patch plan --plan-name database-cluster \
  --target-affinity "REQUIRE nodes(node-role.kubernetes.io/database=true) on node"

# Set power state for all VMs after migration  
kubectl mtv patch plan --plan-name maintenance-migration \
  --target-power-state off

Convertor Pod Optimization Updates

Update convertor pod scheduling for optimal migration performance:

1
2
3
4
5
6
7
8
9
10
11

# Move convertors to high-performance nodes
kubectl mtv patch plan --plan-name data-intensive-migration \
  --convertor-node-selector "node-type=high-io,storage-class=nvme"

# Apply convertor affinity for storage proximity
kubectl mtv patch plan --plan-name storage-migration \
  --convertor-affinity "REQUIRE pods(app=ceph-osd) on node"

# Add labels to convertor pods for monitoring
kubectl mtv patch plan --plan-name monitored-migration \
  --convertor-labels "monitoring=enabled,migration-type=production"

Template and Naming Configuration

Update naming templates for better resource organization:

1
2
3
4
5
6
7
8
9
10
11

# Update PVC naming template
kubectl mtv patch plan --plan-name organized-migration \
  --pvc-name-template "{{.PlanName}}-{{.TargetVmName}}-disk-{{.DiskIndex}}"

# Set volume naming template
kubectl mtv patch plan --plan-name structured-storage \
  --volume-name-template "vol-{{.PVCName}}-{{.VolumeIndex}}"

# Configure network interface naming
kubectl mtv patch plan --plan-name network-organized \
  --network-name-template "{{.TargetVmName}}-{{.NetworkType}}-{{.NetworkIndex}}"

Migration Behavior Configuration

Control various aspects of the migration process:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

# Enable static IP preservation for vSphere VMs
kubectl mtv patch plan --plan-name vsphere-production \
  --preserve-static-ips=true

# Configure shared disk migration
kubectl mtv patch plan --plan-name cluster-workloads \
  --migrate-shared-disks=true

# Enable compatibility mode for older systems
kubectl mtv patch plan --plan-name legacy-systems \
  --use-compatibility-mode=true

# Configure cleanup behavior
kubectl mtv patch plan --plan-name test-migration \
  --delete-guest-conversion-pod=true \
  --delete-vm-on-fail-migration=true

# Skip guest conversion for specific use cases
kubectl mtv patch plan --plan-name raw-disk-migration \
  --skip-guest-conversion=true

# Disable preflight inspection for faster warm migrations
kubectl mtv patch plan --plan-name urgent-migration \
  --run-preflight-inspection=false

Conversion and Guest Conversion Configuration

Configure temporary storage, custom scripts, and virt-v2v image for guest conversion:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

# Set temporary storage for large VM conversion scratch space
kubectl mtv patch plan --plan-name large-migration \
  --conversion-temp-storage-class fast-scratch \
  --conversion-temp-storage-size 500Gi

# Add customization scripts ConfigMap for guest conversion
kubectl mtv patch plan --plan-name windows-migration \
  --customization-scripts migration-scripts/windows-customizations

# Override virt-v2v image for specific migration scenario
kubectl mtv patch plan --plan-name custom-v2v-migration \
  --virt-v2v-image registry.example.com/custom-virt-v2v:latest

# Skip zone node selector for EC2 migrations (EC2 only)
kubectl mtv patch plan --plan-name ec2-migration \
  --skip-zone-node-selector=true

Comprehensive Plan Update Example 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

# Complete plan reconfiguration for production migration
kubectl mtv patch plan --plan-name enterprise-migration \
  --description "Enterprise production migration - Phase 2" \
  --migration-type warm \
  --target-namespace production-v2 \
  --transfer-network production/high-speed-migration \
  --target-labels "environment=production,phase=2,team=platform" \
  --target-node-selector "node-role.kubernetes.io/compute=true" \
  --target-affinity "PREFER nodes(topology.kubernetes.io/zone=us-east-1a) on zone" \
  --target-power-state auto \
  --convertor-labels "migration=production,performance=optimized" \
  --convertor-node-selector "node-type=high-io,network=10gbe" \
  --convertor-affinity "REQUIRE nodes(storage-tier=premium) on node" \
  --preserve-static-ips=true \
  --preserve-cluster-cpu-model=true \
  --pvc-name-template "prod-{{.TargetVmName}}-{{.DiskIndex}}" \
  --delete-guest-conversion-pod=true \
  --run-preflight-inspection=true

How-To: Patching Individual VMs

VM Identity and Configuration

Modify individual VM settings within a plan:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# Change target VM name
kubectl mtv patch planvm --plan-name production-migration --vm-name web-server-01 \
  --target-name web-prod-primary

# Set specific instance type for a VM
kubectl mtv patch planvm --plan-name enterprise-migration --vm-name database-main \
  --instance-type large-memory

# Specify root disk for VMs with multiple disks
kubectl mtv patch planvm --plan-name complex-migration --vm-name multi-disk-vm \
  --root-disk "Hard disk 1"

# Set power state for specific VM
kubectl mtv patch planvm --plan-name maintenance-migration --vm-name critical-db \
  --target-power-state on

VM-Specific Naming Templates

Apply custom naming templates to individual VMs:

1
2
3
4
5
6
7
8
9
10
11

# Custom PVC naming for high-storage VM
kubectl mtv patch planvm --plan-name data-migration --vm-name large-database \
  --pvc-name-template "{{.TargetVmName}}-data-{{.WinDriveLetter}}-{{.DiskIndex}}"

# Custom volume naming for multi-tier application
kubectl mtv patch planvm --plan-name app-migration --vm-name web-tier \
  --volume-name-template "{{.TargetVmName}}-vol-{{.VolumeIndex}}"

# Custom network naming for multi-homed VMs
kubectl mtv patch planvm --plan-name network-migration --vm-name firewall-vm \
  --network-name-template "{{.TargetVmName}}-{{.NetworkType}}-{{.NetworkIndex}}"

Security Configuration

Configure VM-specific security settings:

1
2
3
4
5
6
7
8
9
10
11

# Add LUKS decryption secret for encrypted VM
kubectl mtv patch planvm --plan-name secure-migration --vm-name encrypted-database \
  --luks-secret db-encryption-keys

# Enable NBDE/Clevis passphrase-less disk unlocking via TANG server
kubectl mtv patch planvm --plan-name secure-migration --vm-name nbde-encrypted-vm \
  --nbde-clevis=true

# Override plan-level deletion policy for specific VM
kubectl mtv patch planvm --plan-name test-migration --vm-name experimental-vm \
  --delete-vm-on-fail-migration=true

Adding and Managing Hooks

Attach custom automation to specific VMs:

Adding Hooks 

1
2
3
4
5
6
7
8
9
10
11
12

# Add pre-migration hook to database VM
kubectl mtv patch planvm --plan-name production-migration --vm-name database-primary \
  --add-pre-hook database-backup-hook

# Add post-migration hook to web server
kubectl mtv patch planvm --plan-name production-migration --vm-name web-server-01 \
  --add-post-hook health-check-hook

# Add both pre and post hooks to critical application
kubectl mtv patch planvm --plan-name critical-migration --vm-name app-server-main \
  --add-pre-hook app-quiesce-hook \
  --add-post-hook app-validation-hook

Managing Existing Hooks 

1
2
3
4
5
6
7
8
9
10
11
12
13

# Remove specific hook from VM
kubectl mtv patch planvm --plan-name production-migration --vm-name web-server-01 \
  --remove-hook old-health-check

# Clear all hooks from VM
kubectl mtv patch planvm --plan-name production-migration --vm-name test-vm \
  --clear-hooks

# Replace hook by removing old and adding new
kubectl mtv patch planvm --plan-name production-migration --vm-name database-secondary \
  --remove-hook old-backup-hook
kubectl mtv patch planvm --plan-name production-migration --vm-name database-secondary \
  --add-pre-hook new-enhanced-backup-hook

Comprehensive VM Update Example 

1
2
3
4
5
6
7
8
9
10
11
12
13

# Complete VM customization within plan
kubectl mtv patch planvm --plan-name enterprise-migration --vm-name critical-database \
  --target-name db-prod-primary \
  --instance-type extra-large \
  --root-disk "SCSI disk 1" \
  --target-power-state on \
  --pvc-name-template "{{.TargetVmName}}-{{.WinDriveLetter}}-{{.DiskIndex}}" \
  --volume-name-template "{{.TargetVmName}}-storage-{{.VolumeIndex}}" \
  --network-name-template "{{.TargetVmName}}-net-{{.NetworkIndex}}" \
  --luks-secret database-encryption-secret \
  --delete-vm-on-fail-migration=false \
  --add-pre-hook database-cluster-quiesce \
  --add-post-hook database-cluster-validate

Advanced Patching Scenarios

Scenario 1: Migration Strategy Evolution 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

# Initial plan with basic cold migration
kubectl mtv create plan --name evolving-migration \
  --source vsphere-prod --target openshift-prod \
  --vms "app-server-01,app-server-02,database-01"

# Evolve to warm migration after testing
kubectl mtv patch plan --plan-name evolving-migration \
  --migration-type warm \
  --run-preflight-inspection=true

# Add convertor optimization for warm migration
kubectl mtv patch plan --plan-name evolving-migration \
  --convertor-node-selector "node-type=high-io" \
  --convertor-affinity "REQUIRE nodes(network-speed=10gbe) on node"

# Configure individual database VM for special handling
kubectl mtv patch planvm --plan-name evolving-migration --vm-name database-01 \
  --target-name database-primary \
  --instance-type large-memory \
  --add-pre-hook database-backup-hook \
  --add-post-hook database-validation-hook

Scenario 2: Performance Optimization Through Patching 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

# Start with basic plan
kubectl mtv create plan --name performance-migration \
  --source vsphere-datacenter --target openshift-cluster \
  --vms "where tags.category='performance-critical'"

# Add performance optimizations
kubectl mtv patch plan --plan-name performance-migration \
  --transfer-network performance/dedicated-migration-net \
  --convertor-node-selector "node-type=high-performance,storage=nvme" \
  --convertor-labels "priority=high,performance=optimized" \
  --convertor-affinity "REQUIRE nodes(cpu-type=intel-skylake) on node"

# Optimize target placement
kubectl mtv patch plan --plan-name performance-migration \
  --target-node-selector "node-role.kubernetes.io/compute=true,performance-tier=premium" \
  --target-affinity "PREFER nodes(topology.kubernetes.io/zone=performance-zone) on zone" \
  --target-labels "performance=critical,monitoring=enhanced"

Scenario 3: Incremental Hook Integration 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

# Basic plan without hooks
kubectl mtv create plan --name progressive-automation \
  --source vmware-test --target k8s-test \
  --vms database-test,web-test,cache-test

# Add database-specific automation
kubectl mtv patch planvm --plan-name progressive-automation --vm-name database-test \
  --add-pre-hook database-quiesce \
  --add-post-hook database-health-check

# Add web server automation
kubectl mtv patch planvm --plan-name progressive-automation --vm-name web-test \
  --add-pre-hook lb-drain-connections \
  --add-post-hook web-health-validation

# Add cache server automation
kubectl mtv patch planvm --plan-name progressive-automation --vm-name cache-test \
  --add-post-hook cache-warmup-hook

# Add plan-level notification
# Note: Plan-level hooks require adding to all VMs individually
for vm in database-test web-test cache-test; do
  kubectl mtv patch planvm --plan-name progressive-automation --vm-name $vm \
    --add-post-hook migration-notification
done

Scenario 4: Environment-Specific Customization 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

# Development environment configuration
kubectl mtv patch plan --plan-name dev-migration \
  --target-namespace development \
  --target-labels "environment=dev,auto-shutdown=true" \
  --target-power-state off \
  --delete-vm-on-fail-migration=true \
  --delete-guest-conversion-pod=true

# Production environment configuration  
kubectl mtv patch plan --plan-name prod-migration \
  --target-namespace production \
  --target-labels "environment=prod,backup=required,monitoring=critical" \
  --target-node-selector "node-role.kubernetes.io/production=true" \
  --target-affinity "REQUIRE nodes(reliability-tier=high) on node" \
  --preserve-static-ips=true \
  --preserve-cluster-cpu-model=true \
  --delete-vm-on-fail-migration=false

Patching with Provider Updates

While less common, provider configurations can also be updated:

1
2
3
4
5
6
7
8
9
10
11
12

# Update provider URL (maintenance window)
kubectl mtv patch provider --name my-vsphere-provider \
  --url https://new-vcenter.example.com/sdk

# Update provider credentials
kubectl mtv patch provider --name my-openstack-provider \
  --username new_service_account \
  --password new_secure_password

# Add VDDK image to existing provider
kubectl mtv patch provider --name vsphere-prod \
  --vddk-init-image registry.company.com/vddk:8.0.2

Best Practices: Plan-Level vs. VM-Level Changes

Use Plan-Level Patching When:

  1. Uniform Changes: All VMs need the same configuration updates
  2. Migration Strategy: Changing migration type, network, or target namespace
  3. Performance Optimization: Convertor pod scheduling affects all VMs
  4. Environment Settings: Target cluster configuration applies globally
  5. Template Updates: Naming templates should be consistent across VMs

Use VM-Level Patching When:

  1. Individual Customization: Specific VMs need unique configurations
  2. Selective Hooks: Only certain VMs require automation
  3. Security Requirements: LUKS secrets or power states vary per VM
  4. Instance Types: VMs have different resource requirements
  5. Target Names: Custom naming for specific VMs

Strategic Patching Approach

1. Plan First, Then Customize 

1
2
3
4
5
6
7
8
9
10

# Start with plan-wide optimizations
kubectl mtv patch plan --plan-name enterprise-migration \
  --migration-type warm \
  --target-namespace production \
  --convertor-node-selector "node-type=high-io"

# Then customize specific VMs
kubectl mtv patch planvm --plan-name enterprise-migration --vm-name database-cluster \
  --instance-type extra-large \
  --add-pre-hook cluster-backup

2. Group Similar VMs 

1
2
3
4
5
6
7
8
9
10
11
12
13

# Apply similar configurations to VM groups
for vm in web-01 web-02 web-03; do
  kubectl mtv patch planvm --plan-name web-migration --vm-name $vm \
    --target-labels "tier=web,load-balancer=true" \
    --add-post-hook web-health-check
done

for vm in db-primary db-secondary; do
  kubectl mtv patch planvm --plan-name db-migration --vm-name $vm \
    --instance-type large-memory \
    --add-pre-hook database-backup \
    --add-post-hook database-validation
done

3. Iterative Refinement 

1
2
3
4
5
6
7
8
9
10
11
12
13
14

# Initial basic configuration
kubectl mtv patch plan --plan-name iterative-migration \
  --migration-type cold \
  --target-namespace staging

# Test and refine
kubectl mtv patch plan --plan-name iterative-migration \
  --migration-type warm \
  --convertor-node-selector "storage=ssd"

# Final production configuration
kubectl mtv patch plan --plan-name iterative-migration \
  --target-namespace production \
  --target-affinity "REQUIRE nodes(reliability=high) on node"

Validation and Verification

Verify Plan Changes 

1
2
3
4
5
6
7
8

# Check plan configuration after patching
kubectl get plan production-migration -o yaml

# Verify specific fields
kubectl get plan production-migration -o jsonpath='{.spec.targetNamespace}'

# Check VM configurations
kubectl get plan production-migration -o jsonpath='{.spec.vms[*].name}'

Monitor Patch Results 

1
2
3
4
5
6
7
8

# Watch plan status after patching
kubectl get plan production-migration -w

# Check plan conditions
kubectl describe plan production-migration | grep -A 5 Conditions

# Verify VM-specific changes
kubectl get plan production-migration -o jsonpath='{.spec.vms[?(@.name=="database-01")].hooks}'

Common Patching Scenarios and Solutions

Problem: Migration Too Slow

Solution: Optimize with Patching

1
2
3
4
5
6
7
8

# Add high-performance convertor scheduling
kubectl mtv patch plan --plan-name slow-migration \
  --convertor-node-selector "node-type=high-io,network=10gbe" \
  --convertor-affinity "REQUIRE nodes(storage-tier=premium) on node"

# Switch to warm migration for reduced downtime
kubectl mtv patch plan --plan-name slow-migration \
  --migration-type warm

Problem: VM Naming Conflicts

Solution: Update Naming Templates

1
2
3
4
5
6
7

# Fix naming conflicts with better templates
kubectl mtv patch plan --plan-name naming-conflict \
  --pvc-name-template "{{.PlanName}}-{{.TargetVmName}}-{{.DiskIndex}}"

# Individual VM name fixes
kubectl mtv patch planvm --plan-name naming-conflict --vm-name conflicting-vm \
  --target-name unique-vm-name-prod

Problem: Missing Automation

Solution: Add Hooks Incrementally

1
2
3
4

# Add hooks to specific VMs needing automation
kubectl mtv patch planvm --plan-name manual-migration --vm-name database-server \
  --add-pre-hook backup-automation \
  --add-post-hook validation-automation

Problem: Security Compliance

Solution: Add Security Configurations

1
2
3
4
5
6
7

# Add LUKS support for encrypted VMs
kubectl mtv patch planvm --plan-name secure-migration --vm-name encrypted-vm \
  --luks-secret encryption-keys-secret

# Configure secure target placement
kubectl mtv patch plan --plan-name secure-migration \
  --target-node-selector "security-level=high,compliance=pci-dss"

Next Steps

After mastering plan patching techniques:

  1. Execute Migrations: Learn to start and manage migration execution in Chapter 19: Plan Lifecycle Execution
  2. Handle Problems: Master troubleshooting in Chapter 20: Debugging and Troubleshooting
  3. Optimize Operations: Learn best practices in Chapter 21: Best Practices and Security
  4. AI Integration: Explore advanced automation in Chapter 22: Model Context Protocol (MCP) Server Integration

Previous: Chapter 17: Migration Hooks
Next: Chapter 19: Plan Lifecycle Execution