Junos OS

I'm having real trouble trying to get this to work. I've got a VC consisting of 6 switches, two EX4300 and 4 EX4600s. I suspect drops that I'm seeing across the VC are due to buffer overflows, but am struggling with getting analytics running to get the data to confirm that. 

What I've put in the configuration:

resource-profiles {
    bursts {
        queue-monitoring;
        traffic-monitoring;
        depth-threshold high 999999 low 99;
    }
}
resource {
    system {
        resource-profile bursts;
        polling-interval {
            traffic-monitoring 5;
            queue-monitoring 10;
        }
    }
    interfaces {
        ge-4/0/12 {
            resource-profile bursts;
        }
    }
}
collector {
    local {
        file bursts size 10m files 3;
    }
}

I can't tell if it's because the configuration is bad, or if the feature isn't functional with this setup, but with the above all I get is:

error: peer_daemon: bad daemon: analyticsd

When I run a status on analytics.