DHCP Failover and MCLT configuration implications
  • Updated on 25 Oct 2018
  • 1 minute to read
  • Contributors
  • Print
  • Share
  • Dark
    Light

DHCP Failover and MCLT configuration implications

  • Print
  • Share
  • Dark
    Light

Maximum Client Lead Time (MCLT) is defined as the maximum amount of time that one server can extend a lease for a client's binding beyond the time known by the partner server. It's configured on the primary server of a failover pair and transmitted to the secondary partner when the two are in communication.

It is often a bad idea to make this smaller than the default.

The short explanation is that having a too-small value will cause performance problems in normal operation (but will indeed buy you a quicker recovery if one of the pair fails). Conversely, having a value that is too large will improve performance but means that you have a longer delay at failover.

And this is more detail on why that is: http://tools.ietf.org/html/draft-ietf-dhc-failover-12#section-5.15.

So choosing the right numbers is a calculation based on knowledge of the clients you have and guesstimates on what proportion of the time your servers will be allocating leases to completely new clients versus handling renewals and so on, and then looking at what rate of client requests you think your servers can handle.

Was this article helpful?