On convergence rate of distributed stochastic gradient algorithm for convex optimization with inequality constraints