How the two DHCP failover modes split or reserve addresses, what MCLT does and how relay agents fit in.
From Ultra Transcenders AZ-802 by Tony Rough (coming November 2026)
DHCP failover replicates scopes and leases between two servers so either can serve clients. It is configured per failover relationship and supports DHCPv4 only.
Add-DhcpServerv4Failover; no service restart is needed.| Feature | Load balance mode | Hot standby mode |
|---|---|---|
| Default? | Yes | No |
| Behaviour | Both servers lease simultaneously, split by a hash of the client MAC address | Active server leases; standby leases only if the active server is unavailable |
| Key setting | Load balance percentage, default 50:50 (any ratio 0-100%) | Reserve percentage for the standby, default 5% |
| Best fit | Both servers in the same site | Central standby backing up a server at a remote site |
| Partner lost | Renews partner’s clients for the MCLT; after partner-down state and the MCLT, takes 100% of the pool | Standby uses its reserve for new clients; after the MCLT, can use the whole pool |
The maximum client lead time (MCLT) is the temporary lease duration a server grants when it can’t confirm state with its partner, and the wait before a server in partner-down state takes over the whole pool. The automatic state switchover interval controls moving from communication-interrupted to partner-down.
Add-DhcpServerv4Failover -Name 'DHCP1-DHCP2' -PartnerServer 'dhcp2.example.com' -ScopeId 10.0.0.0 -LoadBalancePercent 50 -SharedSecret 'S3cret!' -MaxClientLeadTime 1:00:00
Invoke-DhcpServerv4FailoverReplication -Name 'DHCP1-DHCP2'DHCP failover works with clustered DHCP: configure it with the cluster’s name or IP address, not a node’s, and replicate any shared secret to all nodes.
Common trap: Changing scope options on one failover partner and assuming the other picks them up - scope changes must be replicated manually, and replication overwrites the partner’s settings.
DHCP discovery uses broadcasts, which routers don’t forward. A DHCP relay agent (an IP helper on a router, or the Remote Access role on Windows Server) forwards client requests to DHCP servers on other subnets. With failover, clients must reach both partners, so configure a relay entry for each server (for example, two ip helper-address lines), or a subnet broadcast address if both servers share a subnet. Router redundancy protocols such as VRRP or HSRP can send duplicate relayed requests and cause inconsistent leases unless configured for DHCP relay.
This note is one section of Ultra Transcenders AZ-802: Administering Windows Server, an independent study guide that explains every topic the exam covers by technology, with comparison tables, diagrams and the common traps, plus a glossary linked to Microsoft Learn.
Due on Amazon in November 2026, in Kindle and paperback editions.
About the book · AZ-802 terms in the glossary · All AZ-802 study notes
Which operations master roles exist per forest and per domain, what each does and the placement guidance for each.
When to deploy an RODC, how staged installation works and how the allowed and denied groups decide which passwords are cached.
Which managed service account type fits a service, its requirements such as the KDS root key, and how Windows Server 2025 delegated MSAs migrate old accounts.
The LSDOU order, which GPO wins a conflict, and how Block Inheritance, Enforced and link order change the result.
What SMB over QUIC needs on the server, the certificate rules, which editions support it and how client access control works.
The Storage Replica topologies, when to use synchronous or asynchronous mode, log volume requirements and the Standard edition limits.
How Windows LAPS rotates and stores local administrator passwords, its policy settings, encryption and the cmdlets to retrieve them.