MEDIUM: server: Implement bounded-load hash algorithm
The consistent hash lookup is done as normal, then if balancing is
enabled, we progress through the hash ring until we find a server that
doesn't have "too much" load. In the case of equal weights for all
servers, the allowed number of requests for a server is either the
floor or the ceil of (num_requests * hash-balance-factor / num_servers);
with unequal weights things are somewhat more complicated, but the
spirit is the same -- a server should not be able to go too far above
(its relative weight times) the average load. Using the hash ring to
make the second/third/etc. choice maintains as much locality as
possible given the load limit.
Signed-off-by: Andrew Rodland <andrewr@vimeo.com>
diff --git a/src/lb_chash.c b/src/lb_chash.c
index a62dfb5..84a2ef3 100644
--- a/src/lb_chash.c
+++ b/src/lb_chash.c
@@ -242,6 +242,34 @@
}
/*
+ * This function implements the "Consistent Hashing with Bounded Loads" algorithm
+ * of Mirrokni, Thorup, and Zadimoghaddam (arxiv:1608.01350), adapted for use with
+ * unequal server weights.
+ */
+int chash_server_is_eligible(struct server *s)
+{
+ /* The total number of slots to allocate is the total number of outstanding requests
+ * (including the one we're about to make) times the load-balance-factor, rounded up.
+ */
+ unsigned tot_slots = ((s->proxy->served + 1) * s->proxy->lbprm.chash.balance_factor + 99) / 100;
+ unsigned slots_per_weight = tot_slots / s->proxy->lbprm.tot_weight;
+ unsigned remainder = tot_slots % s->proxy->lbprm.tot_weight;
+
+ /* Allocate a whole number of slots per weight unit... */
+ unsigned slots = s->eweight * slots_per_weight;
+
+ /* And then distribute the rest among servers proportionally to their weight. */
+ slots += ((s->cumulative_weight + s->eweight) * remainder) / s->proxy->lbprm.tot_weight
+ - (s->cumulative_weight * remainder) / s->proxy->lbprm.tot_weight;
+
+ /* But never leave a server with 0. */
+ if (slots == 0)
+ slots = 1;
+
+ return s->served < slots;
+}
+
+/*
* This function returns the running server from the CHASH tree, which is at
* the closest distance from the value of <hash>. Doing so ensures that even
* with a well imbalanced hash, if some servers are close to each other, they
@@ -254,6 +282,7 @@
struct server *nsrv, *psrv;
struct eb_root *root;
unsigned int dn, dp;
+ int loop;
if (p->srv_act)
root = &p->lbprm.chash.act;
@@ -287,7 +316,23 @@
dp = hash - prev->key;
dn = next->key - hash;
- return (dp <= dn) ? psrv : nsrv;
+ if (dp <= dn) {
+ next = prev;
+ nsrv = psrv;
+ }
+
+ loop = 0;
+ while (p->lbprm.chash.balance_factor && !chash_server_is_eligible(nsrv)) {
+ next = eb32_next(next);
+ if (!next) {
+ next = eb32_first(root);
+ if (++loop > 1) // protection against accidental loop
+ break;
+ }
+ nsrv = eb32_entry(next, struct tree_occ, node)->server;
+ }
+
+ return nsrv;
}
/* Return next server from the CHASH tree in backend <p>. If the tree is empty,