A quorum system is a collection of sets (quorums) every two of which i
ntersect. Quorum systems have been used for many applications in the a
rea of distributed systems, including mutual exclusion, data replicati
on, and dissemination of information. Given a strategy to pick quorums
, the load L (S) is the minimal access probability of the busiest elem
ent, minimizing over the strategies. The capacity S Cap(S) is the high
est quorum accesses rate that can handle, so Cap(S) = 1/L(S). The S av
ailability of a quorum S system L S S is the probability that at least
one quorum survives, assuming that each element fails independently w
ith probability p. A tradeoff between L () and the S availability of i
s shown. We present S four novel constructions of quorum systems, all
featuring optimal or near optimal load, and high availability. The bes
t construction, based on paths in a grid, has a load of O(1/root n), a
nd a failure probability of exp(-Omega(root n)) when the elements fail
with probability p < 1/2. Moreover, even in the presence of faults, w
ith exponentially high probability the load of this system is still O(
1/root n). The analysis of this scheme is based on percolation theory.