job "memstress" { datacenters = ["dc1", "dc2"] type = "batch" group "memstress" { count = 1 restart { mode = "fail" attempts = 0 } reschedule { attempts = 3 interval = "10m" unlimited = false } task "memstress" { driver = "docker" config { image = "progrium/stress" args = [ "-m", "2", "-t", "120" ] } resources { cpu = 4096 memory = 1024 } } } }