We study conditions under which a global error bound in terms of a nat
ural residual exists for a convex inequality system. Specifically, we
obtain an error bound result, which unifies many existing results assu
ming a Slater condition. We also derive two characterizations for a co
nvex inequality system to possess a global error bound; one is in term
s of metric regularity, and the other is in terms of an associated con
vex inequality system. As a consequence, we show that in R-n a global
error bound holds for such a system under the assumption of the zero v
ector in the relative interior of the domain of an associated conjugat
e function along with metric regularity at every point of the feasible
set defined by the system. Finally, we discuss some applications of t
hese results to convex programs.