distributed.jl

using Distributed
# add processes to workspace
addprocs(2)

@everywhere using ModelingToolkit, OrdinaryDiffEq

# create the Lorenz system
@everywhere @parameters t σ ρ β
@everywhere @variables x(t) y(t) z(t)
@everywhere D = Differential(t)

@everywhere eqs = [D(x) ~ σ*(y-x),
       D(y) ~ x*(ρ-z)-y,
       D(z) ~ x*y - β*z]

@everywhere de = ODESystem(eqs)
@everywhere ode_func = ODEFunction(de, [x,y,z], [σ, ρ, β])

@everywhere u0 = [19.,20.,50.]
@everywhere params = [16.,45.92,4]

@everywhere ode_prob = ODEProblem(ode_func, u0, (0., 10.),params)

@everywhere begin

    using OrdinaryDiffEq
    using ModelingToolkit

    function solve_lorenz(ode_problem)
        print(solve(ode_problem,Tsit5()))
    end
end

solve_lorenz(ode_prob)

future = @spawn solve_lorenz(ode_prob)
@test_broken fetch(future)