Site Protection and Disaster Recovery for VMware Cloud Foundation
VCF 9.0 · Pages 83-178 · ID vvs-9x-site-protection-dr
Key Components: VMware Live Recovery, vSphere Replication, NSX, SDDC Manager, vCenter, VCF Operations
External dependencies: This is the VCF 9.0 evolution of the VCF 5.2 Site Protection and Disaster Recovery validated solution. It covers disaster recovery of the VCF management components across two VCF instances: VCF Operations, VCF Operations fleet management, VCF Operations for logs and VCF Operations for networks are replicated and recovered with VMware Live Recovery, while VCF Automation and VCF Identity Broker are recovered by deploying a new instance and restoring a backup (VVS 9.X p.87-88).
External dependencies: Purpose: Instructions for adapting a dual-instance SDDC on top of VCF to provide disaster recovery of SDDC management components using VMware Live Recovery for VCF Automation, VCF Operations, and VCF Operations fleet management.
39 design decisions. The Quality column is the Academy's assessment for defence practice; the VVS documents do not classify decisions by design quality (D133).
| DD-ID | Decision | Quality (Academy view) |
|---|---|---|
| SPR-VLR-CFG-001 | Deploy VLR instance in protected and recovery VCF instances. | Recoverability |
Decision: Deploy VLR instance in protected and recovery VCF instances. Rationale: Allows orchestrated recovery of VCF mgmt components in another VCF instance. Implication: No significant trade-offs identified for this decision. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-002 | Deploy each VLR in the management domain. | Manageability |
Decision: Deploy each VLR in the management domain. Rationale: Consistent deployment model for all mgmt applications. Implication: No significant trade-offs identified for this decision. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-003 | Deploy the VMware Live Recovery virtual appliance using the Light deployment type. | RecoverabilityManageability |
Decision: Deploy the VMware Live Recovery virtual appliance using the Light deployment type. Rationale: Provides the highest level of availability by protecting the management components. The Light size accommodates the number of protected management VMs (Management workloads with failover support), three protection groups and three recovery plans (VVS 9.X Table 100). Implication: None. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-004 | Use vSphere Replication in VLR as VM replication protection. | Manageability |
Decision: Use vSphere Replication in VLR as VM replication protection. Rationale: Flexibility in storage/vendor; minimizes SRA compatibility overhead. Implication: All mgmt components must be in same cluster; reduced total VM count vs storage-based replication. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-005 | Use VLR to automate recovery of: VCF Operations, VCF Operations fleet mgmt, VCF Operations for… | ManageabilityRecoverability |
Decision: Use VLR to automate recovery of: VCF Operations, VCF Operations fleet mgmt, VCF Operations for logs, VCF Operations for networks. Rationale: Automated runbook; RTO <= 4 hours. Implication: No significant trade-offs identified for this decision. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-006 | Configure PIT instances keeping 3 copies over 24-hour period. | Recoverability |
Decision: Configure PIT instances keeping 3 copies over 24-hour period. Rationale: Application integrity after DR recovery event. Implication: More PIT = more disk usage on vSAN datastore. Component: VMware Live Recovery | ||
| SPR-VLR-NET-001 | Place VLR on the management network. | Manageability |
Decision: Place VLR on the management network. Rationale: Co-locate with VCF components. Implication: No significant trade-offs identified for this decision. Component: VMware Live Recovery | ||
| SPR-VLR-NET-002 | Allocate static IP to VLR instances. | Manageability |
Decision: Allocate static IP to VLR instances. Rationale: Removes DHCP constraints/risks on management networks. Implication: Precise IP management. Component: VMware Live Recovery | ||
| SPR-VLR-NET-003 | Configure A + PTR DNS records for each VLR. | ManageabilitySecurity |
Decision: Configure A + PTR DNS records for each VLR. Rationale: FQDN accessibility. Implication: DNS infrastructure required; manage DNS records; firewalls allow DNS. Component: VMware Live Recovery | ||
| SPR-VLR-NET-004 | Configure VLR to use NTP servers (not VMTools). | Security |
Decision: Configure VLR to use NTP servers (not VMTools). Rationale: Accurate time sync; prevent mismatch. Implication: NTP services required; firewalls allow NTP. Component: VMware Live Recovery | ||
| SPR-VLR-NET-005 | Use dedicated vDS portgroup for VLR traffic. | Security |
Decision: Use dedicated vDS portgroup for VLR traffic. Rationale: Isolates VLR traffic from other system traffic. Implication: Manually create portgroup. Component: VMware Live Recovery | ||
| SPR-VLR-NET-006 | Use dedicated VLAN for VLR traffic. | Security |
Decision: Use dedicated VLAN for VLR traffic. Rationale: Traffic isolation. Implication: Manually create VLAN. Component: VMware Live Recovery | ||
| SPR-VLR-NET-007 | Use dedicated VMkernel interface on each ESX for VLR traffic. | Security |
Decision: Use dedicated VMkernel interface on each ESX for VLR traffic. Rationale: Traffic isolation. Implication: Manually create VMkernel interface. Component: VMware Live Recovery | ||
| SPR-VLR-NET-008 | Assign static IP to dedicated VMkernel for VLR traffic. | Manageability |
Decision: Assign static IP to dedicated VMkernel for VLR traffic. Rationale: Removes DHCP risks. Implication: Precise IP management. Component: VMware Live Recovery | ||
| SPR-VLR-LCM-001 | Lifecycle manage VLR via native appliance tools. | Manageability |
Decision: Lifecycle manage VLR via native appliance tools. Rationale: VLR not managed by SDDC Manager. Implication: Deployment, patching, updates, upgrades without native automation. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-001 | Configure service account in each vCenter for app-to-app communication VLR > vSphere, in SSO… | Manageability |
Decision: Configure service account in each vCenter for app-to-app communication VLR > vSphere, in SSO Administrators group. Rationale: VLSR DR orchestration + site pairing; VR replication; accountability; limited blast radius. Implication: Maintain service account lifecycle outside VCF. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-003 | Configure password expiration policy for VLR appliance. | Manageability |
Decision: Configure password expiration policy for VLR appliance. Rationale: Align with org/compliance policies (local users only). Implication: Manage via appliance console or SSH. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-004 | Configure password complexity policy for VLR appliance. | Manageability |
Decision: Configure password complexity policy for VLR appliance. Rationale: Align with org/compliance policies (local users only). Implication: Manage via console or SSH. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-005 | Configure account lockout policy for VLR appliance. | Manageability |
Decision: Configure account lockout policy for VLR appliance. Rationale: Align with org/compliance policies (local users only). Implication: Manage via console or SSH. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-006 | Change VLR root password on recurring/event-initiated schedule. | Manageability |
Decision: Change VLR root password on recurring/event-initiated schedule. Rationale: Root passwords never expire by default. Implication: Routine password changes required. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-007 | Change VLR admin password on recurring/event-initiated schedule. | Manageability |
Decision: Change VLR admin password on recurring/event-initiated schedule. Rationale: Admin passwords never expire by default. Implication: Routine password changes required. Component: VMware Live Recovery | ||
| SPR-VLR-SEC-008 | Replace default self-signed cert with CA-signed in each VLR. | Security |
Decision: Replace default self-signed cert with CA-signed in each VLR. Rationale: All externally facing Web UI + cross-product communication encrypted. Implication: Must have PKI access. Component: VMware Live Recovery | ||
| SPR-VLR-RP-001 | Use prioritized startup order for VCF Operations + VCF Operations FM appliance nodes. | ManageabilityRecoverability |
Decision: Use prioritized startup order for VCF Operations + VCF Operations FM appliance nodes. Rationale: Individual VCF Ops nodes started in order for operational monitoring. Implication: Maintain customized recovery plan if node count increases. Component: VMware Live Recovery | ||
| SPR-VLR-RP-002 | Use prioritized startup order for VCF Operations for logs nodes. | ManageabilityRecoverability |
Decision: Use prioritized startup order for VCF Operations for logs nodes. Rationale: Cloud automation services restored; restored within 4h. Implication: VMware Tools on each VCF Ops Logs node. Component: VMware Live Recovery | ||
| SPR-VLR-RP-003 | Use prioritized startup order for VCF Operations for networks nodes. | ManageabilityRecoverability |
Decision: Use prioritized startup order for VCF Operations for networks nodes. Rationale: Cloud automation services restored; restored within 4h. Implication: VMware Tools on each VCF Ops Networks node. Component: VMware Live Recovery | ||
| SPR-VLR-RP-005 | Do not run test recovery of recovery plans. | Recoverability |
Decision: Do not run test recovery of recovery plans. Rationale: Because some of the protected applications may use a load balancer, it is not possible to bring the applications online in an isolated test network. DNS resolution is also unavailable in an isolated test network (VVS 9.X Table 116). Implication: You cannot test disaster recovery without impacting the running production applications. Component: VMware Live Recovery | ||
| SPR-DNS-NET-001 | With multiple VCF instances, configure DNS settings for each protected component to use DNS servers… | Availability |
Decision: With multiple VCF instances, configure DNS settings for each protected component to use DNS servers across all VCF instances. Rationale: Higher DNS availability and resilience during DR/migration. Implication: As you scale from single to multiple instances, update DNS settings on each protected component. Component: DNS | ||
| SPR-NTP-NET-001 | With multiple VCF instances, configure NTP settings across all instances. | Availability |
Decision: With multiple VCF instances, configure NTP settings across all instances. Rationale: Higher NTP availability during DR/migration. Implication: Update NTP settings as instances scale. Component: NTP | ||
| SPR-VCFO-CFG-001 | Install VLR management pack for VCF Operations. | Manageability |
Decision: Install VLR management pack for VCF Operations. Rationale: Establishes communication between VCF Ops and VLR endpoints. Implication: Manual management pack installation. Component: VCFO | ||
| SPR-VCFO-CFG-002 | Configure the VMware Live Recovery endpoints — VMware Live Site Recovery and vSphere Replication… | Manageability |
Decision: Configure the VMware Live Recovery endpoints — VMware Live Site Recovery and vSphere Replication — to use the VCF Operations collector or collector group for the VCF Instance. Rationale: Offloads data collection from analytics cluster. Implication: No significant trade-offs identified for this decision. Component: VCFO | ||
| SPR-VLR-LOG-001 | When using VLR, install VCF Operations for logs agent on VLR appliance. | Manageability |
Decision: When using VLR, install VCF Operations for logs agent on VLR appliance. Rationale: Simplifies configuration via agent-based log forwarding. Implication: Must configure VCF Ops for logs as aggregation target. Component: VMware Live Recovery | ||
| SPR-VLR-LOG-002 | Configure the VCF Operations for logs agent to transmit logs from the VMware Live Recovery instance… | Manageability |
Decision: Configure the VCF Operations for logs agent to transmit logs from the VMware Live Recovery instance to the adjacent VCF Operations for logs in the VCF instance using the ingestion API, cfapi, on port 9000. Rationale: Ensures transmission of logs to centralized system. Implication: This configuration is unencrypted. To encrypt the log transmission with TLS, update the agent configuration on the VMware Live Recovery instance (/etc/liagent.ini) to send logs by the cfapi ingestion API on port 9543 (VVS 9.X p.165-166). Component: VMware Live Recovery | ||
| SPR-VR-CFG-001 | Deploy each VMware Live Recovery appliance in the vCenter it will be registered with. | Manageability |
Decision: Deploy each VMware Live Recovery appliance in the vCenter it will be registered with. Rationale: VMware Live Recovery must be deployed in the vCenter it is registered with, as it discovers the certificate thumbprint during the OVF deployment. Implication: None. Component: VMware Live Recovery | ||
| SPR-VR-CFG-002 | Deploy each VMware Live Recovery appliance using the 4 vCPU size. | Performance |
Decision: Deploy each VMware Live Recovery appliance using the 4 vCPU size. Rationale: Accommodates the replication of the expected number of virtual machines of the protected components (VCF Operations, VCF Operations fleet management, VCF Operations for logs, VCF Operations for networks). Implication: None. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-003-b | Do not activate guest OS quiescing in the policies for the management virtual machines in VMware… | RecoverabilityAvailability |
Decision: Do not activate guest OS quiescing in the policies for the management virtual machines in VMware Live Recovery. Rationale: Not all management virtual machines support guest OS quiescing; using it might result in an outage. (The source reuses SPR-VLR-CFG-003 for this replication-policy decision; the -b suffix is added here, VVS 9.X Table 114, p.172-173.) Implication: The replicas stored in the target VCF instance are crash-consistent rather than application-consistent. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-004-b | Activate network compression on the management virtual machine policies in VMware Live Recovery. | Performance |
Decision: Activate network compression on the management virtual machine policies in VMware Live Recovery. Rationale: Reduces the footprint of the replication traffic on the network and the buffer memory used on the appliance. (The source reuses SPR-VLR-CFG-004 for this replication-policy decision; the -b suffix is added here, VVS 9.X Table 114, p.172-173.) Implication: Compression and decompression require CPU resources on the source site as more virtual machines are protected. Component: VMware Live Recovery | ||
| SPR-VLR-CFG-005-b | Configure a recovery point objective (RPO) of 15 minutes on the management virtual machine policies… | Recoverability |
Decision: Configure a recovery point objective (RPO) of 15 minutes on the management virtual machine policies in VMware Live Recovery. Rationale: Ensures that a management application failing over after a recovery event contains all data except the changes of the last 15 minutes. (The source reuses SPR-VLR-CFG-005 for this replication-policy decision; the -b suffix is added here, VVS 9.X Table 114, p.172-173.) Implication: Any changes made in the 15 minutes before a disaster recovery event are lost. Component: VMware Live Recovery | ||
| SPR-VCFOPS-SEC-001 | Configure a service account in vCenter with global permissions for application-to-application… | SecurityManageability |
Decision: Configure a service account in vCenter with global permissions for application-to-application communication from the VMware Live Recovery adapters in VCF Operations to vSphere and VMware Live Recovery. Rationale: The adapters access vSphere and VMware Live Recovery with the minimum set of permissions required to collect metrics; a compromised account has restricted reach and request tracking is accountable. Implication: You must maintain the life cycle and availability of the service account outside of the SDDC stack. Component: VCF Operations | ||
| SPR-VCFOPS-SEC-002 | Configure a service account in each vCenter with global permission for application-to-application… | SecurityManageability |
Decision: Configure a service account in each vCenter with global permission for application-to-application communication from the VCF Operations adapters to vSphere, and assign the VRM replication viewer role. Rationale: The adapters access vSphere and VMware Live Recovery with the minimum set of permissions required to collect metrics. Implication: You must maintain the life cycle and availability of the service account outside of the SDDC stack. Component: VCF Operations | ||
Prerequisites
- VCF management domain operational
- DNS/NTP configured
- CA infrastructure in place
Implementation Procedure
Implementation
Environment: VCF version in support matrix; per Before You Apply; Site Protection and Disaster Recovery tab of Planning Workbook; VCF healthy.
DNS: Required DNS entries in forward and reverse zones
Network: IP mobility between VCF instances; jumbo frames; L3 routing; max 150ms latency; bandwidth sized with VLR Calculator
Software: VLR .iso mounted
License: VLR license with quantity per design
Active Directory: AD Domain Controllers, service accounts, security groups
CA: CA available
Reconfigure DNS NTP On VCF Management Components
VCF Operations Fleet Mgmt Appliance
DNS: SSH to fleet mgmt node as root. Edit /etc/systemd/network/10-eth0.network - add DNS entry. Edit /etc/systemd/resolved.conf - modify Domains entry (space separated).
NTP: VCF Ops interface > Fleet Management > Lifecycle > Settings > NTP Servers > Add NTP Server or Edit Server Selection
VCF Operations Nodes
DNS: SSH to primary node as root. Edit /etc/systemd/network/10-eth0.network and /etc/systemd/resolved.conf (same as fleet mgmt). Repeat for all cluster nodes + VCF Ops Logs + VCF Ops Networks.
NTP: VCF Ops > Administration > Control Panel > Cluster Management > Actions > Network Time Protocol Settings > Add NTP
VCF Operations For Logs
NTP: SSH to primary logs node as root. Edit /etc/ntp.conf - add 'server ntp.lax.rainpole.io iburst'. Repeat for each.
VCF Operations For Networks
NTP: SSH to primary networks node as support user. sudo vi /etc/ntp.conf - add server. Repeat for each.
VCF Automation
DNS: Take backup first (per VCF Fleet Management). VCF Ops > Fleet Management > Lifecycle > VCF Management > Components > automation > ellipsis > Update DNS Configuration > Add Server, Change Priority, Run Precheck, Finish.
NTP: Same path > Update NTP Configuration > Add, Priority, Precheck, Finish
VCF Identity Broker
DNS: Take backup first. VCF Ops > Fleet Management > Lifecycle > VCF Management > Components > identity broker > ellipsis > Update DNS Configuration
NTP: Same > Update NTP Configuration
Deploy VLR Appliance
- Method: Deploy OVF template in vSphere Client for mgmt domain
- Placement: Default mgmt cluster
- Storage: Mgmt domain vSAN datastore
- Network: Management network
Day-2 Operations Tasks
Verify the site pairing
After installation, outage or certificate changeProtected-instance vCenter > Menu > Site recovery: VMware Live Site Recovery and vSphere Replication status must be OK (green) for both instances. If not, check connectivity between the vCenter instances and the VLR appliances and re-accept certificates.
Operational verification after an outage
After any VLR or vCenter outagePower on the VLR appliance from vCenter, wait for the services, then confirm vSphere Replication and VMware Live Site Recovery are operational on the Site recovery page and that the recovery plans show Ready.
Planned migration or disaster recovery of VCF Operations
Planned maintenance or DRRun the VCF Operations recovery plan from the protected (or, in DR, the recovery) instance. Priority order per SPR-VLR-RP-001: fleet management appliance first, then primary, replica, data nodes. Answer the prompts: verify the load balancer VIP responds to ping, power off remote collectors. Then reconfigure DNS/NTP if needed and verify the analytics cluster nodes and collectors in the VCF Operations admin UI.
Recovery plan for VCF Operations for logs and for networks
Planned maintenance or DRSame model: primary node priority 1 (Highest) for Operations for logs; platform node 1 priority 1 for Operations for networks; add user prompts before the next priority group to verify dependencies.
NSX load balancer and Tier-1 reconfiguration at failover
Every failover and failbackMove the load balancer to the standby Tier-1 on the cross-instance segment in the recovery instance (active-standby), wait for Success, and confirm dynamic routing on the Tier-1 so the recovered VIPs are advertised.
Reprotect and failback
After a completed failoverReprotect the recovery plan (right-click > Reprotect, VVS 9.X p.149) to reverse replication (irreversible step), re-verify the pairing, then run the plan again as a planned migration to fail back once the original instance is healthy.
Certificate management
Before expiry or on compromiseGenerate signed certificates for each VLR appliance; in a multi-instance environment replace them and update the connected components (site pairing, vCenter registration) so the pairing stays trusted; re-verify.
Password management and expiration policy
Per policySSH to the VLR appliance as admin and switch to root; set the expiry with chage --maxdays <value> root and repeat for the admin user; repeat on the recovery instance. Rotate the vCenter service accounts (SPR-VLR-SEC-* and SPR-VCFOPS-SEC-*) outside SDDC Manager.
Monitoring and alerting
ContinuousVCF Operations with the VMware Live Recovery adapter: validate the connection, accept the certificate, confirm Collecting; monitor analytics-cluster node status and collector status; alert on replication RPO violations.
Life cycle management
Per releasePatch and upgrade the VLR appliances with their native tools in both instances (SPR-VLR-LCM-001); keep versions aligned with the VCF BOM; SDDC Manager does not manage VLR.
Monitoring Points
- Verify Pairing
- Check: VMware Live Site Recovery + vSphere Replication status = OK (green)
- Power on VLR via vCenter. Verify VR + VLSR operational after startup.
- Monitoring And Alerting
- Validate connection, accept cert, verify Collecting status
- Checklist
- Activation Assessment: Verify DR is required (e.g., extended instance outage); plan for business continuity
- When prompted, verify LB VIP responds to ping, Dismiss
- Monitor analytics cluster node status (Not Running/Offline)
- Monitor all collectors until Not Running/Offline
Troubleshooting
Likely Panelist Questions
Q: Why the Light deployment type for VMware Live Recovery?
SPR-VLR-CFG-003 (Table 100): Light accommodates the protected management VMs, three protection groups and three recovery plans while giving the highest availability for the management components; larger sizes buy nothing for this scope.
Q: How do you validate DR if test recovery is not run?
SPR-VLR-RP-005: applications behind a load balancer cannot start in an isolated test network and DNS is absent there, so DR is validated by planned migrations in maintenance windows; the accepted implication is that DR cannot be tested without touching production.
Q: Why 15-minute RPO and no guest quiescing?
15 minutes bounds data loss for the management applications; quiescing is off because not every management VM supports it and it can cause an outage — replicas are crash-consistent (SPR-VLR-CFG-003/005 replication-policy rows).
Q: Why does the fleet management appliance start first?
SPR-VLR-RP-001 and p.119: the fleet management appliance (priority 1), then primary, replica and data nodes; collectors are site-local and are powered off by a plan prompt.
Failure Scenarios
Trade-off Analysis
IP mobility: NSX Federation stretched segments vs stretched L2
Chosen: Cross-instance NSX segment with a standby Tier-1 and load balancer in the recovery instance
Justification: Keeps management IPs through failover without stretched physical L2; cost is an NSX Federation design and manual load balancer attach/detach.
Testability vs load-balanced applications
Chosen: No test recovery; planned migrations validate the plans (SPR-VLR-RP-005)
Justification: Isolated test networks cannot host load-balanced applications or DNS.
Appliance size
Chosen: Light VLR deployment type, 4 vCPU appliance (SPR-VLR-CFG-003, SPR-VR-CFG-002)
Justification: Matches the protected management footprint; larger tiers add resources without benefit.
Quiz — Site Protection & DR (VVS 9.x / VCF 9.0)
- 24 hours
- 8 hours
- 4 hours
- 1 hour
- 24 hours
- 15 minutes
- 1 hour
- 4 hours
- VCF Operations stack (Ops, Ops FM, Ops Logs, Ops Networks)
- vCenter + SDDC Manager
- VCF Identity Broker + VCF Automation
- SDDC Manager + NSX
- 3 over 24 hours
- 10
- 200
- 1
- Static routing only
- BGP without overlay
- DHCP
- NSX Federation with stretched overlay segments OR stretched L2 networks
- VCF Operations for networks
- VCF Automation
- VCF Operations for logs
- VCF Operations
- Clustered applications such as VCF Operations cannot run in a test bubble
- Test recovery breaks the replication relationship
- Protected applications behind a load balancer cannot come online in the isolated test network and DNS is unavailable there
- Test recovery is unsupported with vSphere Replication
- No change needed
- Delete and recreate
- Protected -> Secondary, Recovery -> Primary
- Both Primary
- All components simultaneously
- Identity Broker/Automation, then Ops Networks, then Ops Logs, then take Ops cluster offline and shut down
- VCF Ops first
- Random order
- Take cluster offline via admin UI System status
- Snapshot each node
- Force shutdown
- Nothing
- 50 ms
- 150 ms
- 100 ms
- 300 ms
- Force-provisioning
- Nothing special
- Disable SPBM
- Maximum redundancy
- Read-Only
- Administrators
- Custom
- No Access
- Test recovery
- Reprotect
- Planned migration
- Disaster recovery
- To update dashboards
- To enable backups
- To enable day-2 operations (power on/off VMs) after extended stay in recovery - so VCF Ops FM knows new VM location
- Required by license
Flashcards — Site Protection & DR (VVS 9.x / VCF 9.0)
Labs
Lab 1: Deploy VLR and Pair Across Two VCF Instances
Deploy VLR appliances in both VCF instances, reconfigure DNS/NTP across instances, and pair VLR via a cross-instance site pair.
Starting State: Two VCF 9.0 instances operational; IP mobility in place between the instances — here NSX Federation global overlay networks (stretched L2 is the other documented option, VVS 9.X p.87); DNS/NTP/CA services in place; VLR .iso mounted.
- Reconfigure DNS/NTP on VCF Ops FM, VCF Ops, VCF Ops Logs, VCF Ops Networks, VCF Automation, VCF Identity Broker to use both instances' DNS/NTP servers (SSH edits + VCF Ops UI updates)
- Create dedicated VDS portgroup for VLR traffic in each instance (Static binding, Elastic, Route based on physical NIC load)
- Create dedicated VLAN for VLR traffic; deploy VMkernel on each mgmt cluster ESX host (vSphere Replication + NFC, static IPv4)
- Deploy VLR OVF template in each instance (mgmt domain, vSAN, mgmt network, SSHD enabled, ipv4 static)
- Power on VLR; browse to https://<vlr_fqdn>:5480 as admin
- Replace default cert: Generate CSR, submit to AD CS, upload signed cert
- Create vCenter SSO service account in each instance (Admin group member)
- Register VLR with vCenter SSO: PSC port 443, select vCenter, Extension ID com.vmware.vcDr
- In protected vCenter > VMware Live Site Recovery > Open > New site pair
- Pair type: 'Pair with peer vCenter Server in different SSO domain'; select services VLSR + VR
- Verify site pair status = OK (green)
Lab 2: Configure VCF Operations Recovery Plan and Run Planned Migration
Configure VM folder/network/resource mappings, set up VR protection for VCF Operations stack with RPO 15min + 3 PIT, create a recovery plan with prioritized startup, then run a planned migration and reprotect.
Starting State: Lab 1 complete. VCF Operations cluster deployed in protected instance. AD DC + DNS/NTP/CA + SMTP + Syslog services available at both sites. NSX Edge cluster configured in both instances for North-South routing. NSX load balancer deployed/configured in both.
- In protected VLR site pair > View details > Configure > Network mappings > New (manual) > select port groups for mgmt network
- Folder mappings > New > manual > select 'VCF Operations' VM folders
- Resource mappings > New > select mgmt cluster in both instances
- VLR /dr > Replications > Outgoing > New
- Select VCF Ops cluster VMs; Target datastore: per workbook (Disk format 'Same as source', VM storage policy 'vSAN Default Storage Policy' — VVS 9.X p.118)
- Replication settings: RPO 15 minutes, PIT 3 copies / 24 hours, Network compression Selected, Guest OS quiescing NOT selected
- Protection group: Add to new group 'vcf-ops-pg'
- Recovery plan: Add to new plan 'vcf-ops-rp'
- In the recovery plan (Recovery steps tab), set the startup priority: Priority group 1 (Highest) VCF Operations fleet management appliance, 2 VCF Operations primary node, 3 replica node, 4 data node(s). Add a prompt before priority 2 to verify the load balancer, and a prompt to power off all running VCF Operations remote collectors (collectors are site-local and are not recovered) — VVS 9.X p.119.
- Repeat for VCF Ops Logs, VCF Ops Networks in separate recovery plans with VMware Tools startup order
- Before planned migration: Shut down Identity Broker + Automation (Fleet Mgmt Power Off), VCF Ops Networks, VCF Ops Logs, VCF Ops (cluster offline first)
- NSX Federation tasks: reconfigure Tier-1 Gateway (protected=Secondary, recovery=Primary); disconnect protected LB; connect recovery LB
- VLR /dr in protected > site pair > Recovery plans > vcf-ops-rp > Recovery steps > Run
- Confirmation: 'I understand...' + Recovery type: Planned migration; Finish
- When prompted, verify LB VIP responds to ping; dismiss
- Power off protected collectors, monitor nodes Offline, power on recovery collectors, bring cluster online
- Verify recovery plan completes successfully
- After extended operation in recovery, Reprotect: right-click plan > Reprotect > confirm irreversible > Finish
- Monitor Reprotect status = Ready
Lab 3: Perform VCF Identity Broker and VCF Automation Recovery via Backup/Restore
Take backups of VCF Identity Broker and VCF Automation, simulate DR by powering off in protected, deploy new instances in recovery, and restore from backup.
Starting State: VCF Identity Broker + VCF Automation deployed in protected VCF instance. VCF Fleet Management configured with SFTP backup (HA between sites or replicated). Known-good backups exist.
- Verify backup of VCF Identity Broker via VCF Fleet Management (VCF Ops > Fleet Mgmt > Lifecycle > Components > identity broker > Backup and Restore > Backup)
- Verify backup of VCF Automation similarly
- Note backup path: vcf/backups/cluster-name/version/component-name/timestamp
- IMPORTANT: Do not modify VCF SSO or VCF Automation config during failover (changes will be lost)
- Use local VCF Operations user (VCF SSO-based logins won't work while Identity Broker offline)
- Power off VCF Identity Broker in protected: Fleet Mgmt > Components > identity broker > Power Off > Submit
- Power off VCF Automation similarly
- In recovery VCF Ops: Fleet Mgmt > Lifecycle > Components > Add Component > identity broker
- Deployment: New Install; Certificate: select Identity Broker cert; Infrastructure: recovery mgmt vCenter; Network: per workbook; Components: per workbook
- Run Precheck; Submit; monitor task
- Once deployed, Components > identity-broker (ensure recovery instance) > Backup and Restore > Restore
- Select 'Restore from the latest backup on the cluster' OR pass backup path
- Restore; monitor Task pane
- Repeat full procedure for VCF Automation (use same deployment type as protected)
- After DR validation, reverse: redeploy + restore VCF Identity Broker + Automation in protected when available