Plan and leases
- ISC dhcpd
- conf + lease file
- Lattice
- One catalog
Replace
Replace ISC dhcpd
Keep BIND.
Replace ISC dhcpd: subnet, range, and host in conf — plus dhcpd.leases — become rows dhcpd reads. One catalog. Search the MAC instead of grepping the lease file. Discover on A and Ack on B is the happy path.
On-prem · Ubuntu tarball · Community: 1,000 managed IPs · No phone-home
Beta signup issues a lifetime perpetual Starter license so you can test three-node HA. Not a sale.
Today — two systems
Plan
- Spreadsheet
- dhcpd.conf
DHCP
- MMC scopes
- Lease file
Keep them in sync yourself.
Lattice — one catalog
The reservation you save is the next Offer.
- prefix
- subnet
- pool
- reservation
- assignment
- lease
dhcpd serves these rows. Console writes them. There is no second copy to sync. A/PTR follow the grant into the DNS you already run.
Cutover
Lattice does not ingest dhcpd.conf or dhcpd.leases. Read subnet, range, and host { hardware; fixed-address }, map them, POST /api/v1. Small dnsmasq files: enter them in Addresses. Shared-network and allow/deny pools are a human pass.
Playbook: Migrate · HA is not Microsoft failover — three-node VIP.
A Tuesday after cutover
ISC keeps a plan in conf and reality in the lease file. Lattice has one catalog, so Work can name the failure.
9:05
What IP does this MAC have?
Helpdesk pastes a suffix. Search lands on the lease. Neteng is not the lookup API.
11:20
Give the printer .10 forever
The reservation you save is the next Offer. No scope to edit, no sheet to sync.
2:40
Laptop will not get an address
Work names the queue. No capture first. There is no primary to restart.
4:10
Who had this IP last Tuesday?
Occupancy is every remaining lease row. As-of a day. CSV. Not Audit, not a ping.
Scale-out / failover
- ISC dhcpd
- Peer / split pools
- Lattice
- Same pool on every node (Starter: extra dhcpd + three-node VIP). Beta signup includes lifetime Starter.
Find a MAC
- ISC dhcpd
- grep dhcpd.leases
- Lattice
- Console search
Who held this IP
- ISC dhcpd
- grep dhcpd.leases
DNS
- ISC dhcpd
- Scripts / nsupdate / BYO
- Lattice
- A/PTR after Ack into DNS you keep
Cutover
- ISC dhcpd
- Read conf host/subnet/range → map → API (no conf/lease ingest)
- Lattice
- Dump → map → API
What hurts in ISC dhcpd
conf is the plan, the lease file is reality
A host block is intent. dhcpd.leases is what happened. They drift. grep is the helpdesk.
Failover is a peer
Split pools, primary/secondary. Request on the node that did not Offer is a special case. Lattice is the opposite: that is normal.
DNS is a script
nsupdate on a hook, or you do not do it. Ack should not wait either way — but the plan and the name should not live in two files.
What Lattice does
No dhcpd.conf for pools
Pools, reservations, and excludes are IPAM rows. Every dhcpd process loads the same catalog.
Search the MAC
Console search lands on the reservation or the lease. Cisco dotted MAC works.
Who held this IP
Occupancy is every remaining lease row. As-of a day. CSV. Not grep dhcpd.leases.
Keep BIND
TSIG after Ack into the zones you already run. Option 6 stays on BIND or Unbound. lattice-dns is optional, not required.
Same pool, any node
Community is one Ubuntu node. Same binaries. A Starter file adds dhcpd processes on the same database (Discover on one node, Ack on another, same pool) and three-node HA behind a VIP. Signup includes a lifetime perpetual Starter license so you can test HA.
FAQ
- Will Lattice ingest my dhcpd.conf?
- No. Read subnet, range, and host blocks, map them, POST /api/v1. dhcpd.leases is not an import file. Read the migrate playbook.
- Do I have to move DNS off BIND?
- No. Keep BIND. TSIG after Ack into the zones you already run. Option 6 stays on BIND or Unbound.
- Is Lattice an ISC failover peer?
- No. There is no failover protocol. Paid HA is three Ubuntu VMs, one Postgres, one VIP. Discover on one node and Ack on another is the same pool. Read the HA docs.