Struct Rendezvous 

pub struct Rendezvous { /* private fields */ }

Weighted Rendezvous (Highest Random Weight) hashing.

For an affinity key, the chosen backend is the one maximizing a stable, per-(key, backend) score. With no key it degrades to plain round-robin; every datapath supplies one for a client it can identify (the UDP affinity key, and the client key of HTTP, HTTPS and TCP, sozu-proxy/sozu#524), so the fallback is reached only for a request with no source address.

Weighting

We use the standard continuous weighted-rendezvous score (Schindelhauer–Schomaker / “logarithmic method”):

    score(key, backend) = -weight / ln(h)

where h = hash64(seed, key, backend_addr) / 2^64 ∈ (0, 1). Because ln(h) < 0, the score is positive and grows with weight, so heavier backends win proportionally more keys while the choice stays deterministic and minimally disruptive: removing a non-winning backend cannot change the winner, and adding one only steals the keys for which it now scores highest. Selection is O(N) per key.

Implementations

impl Rendezvous

pub fn new() -> Self

pub fn with_seed(seed: u64) -> Self

Trait Implementations

impl Debug for Rendezvous

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for Rendezvous

fn default() -> Self

Returns the “default value” for a type.

impl LoadBalancingAlgorithm for Rendezvous

fn next_available_backend( &mut self, key: Option<u64>, backends: Candidates<'_>, ) -> Option<Rc<RefCell<Backend>>>

Select the next backend among candidates.

fn rebuild(&mut self, backends: &[Rc<RefCell<Backend>>])

Called by the control plane when the live backend set for a cluster changes, so table-based policies (Maglev) can recompute their lookup table off the datapath. The default is a no-op; Maglev rebuilds its table, and RoundRobin (with the round-robin fallback of Rendezvous and Maglev) resets its running weights.