MongoDB slow log (profiler) shows many many "killcursors" action when heavy load comes, why? and what killcursors do?

Question

2015 - 05 - 13 T21: 55: 10.876 + 0800 I QUERY[conn8674] killcursors keyUpdates: 0 writeConflicts: 0 numYields: 0 locks: {
    Global: {
    acquireCount: {
        r: 1
    }
    },
    MMAPV1Journal: {
    acquireCount: {
        r: 1
    }
    },
    Database: {
    acquireCount: {
        r: 1
    }
    },
    Collection: {
    acquireCount: {
        R: 1
    }
    }
}
1641 ms 2015 - 05 - 13 T21: 55: 10.877 + 0800 I QUERY[conn8618] killcursors keyUpdates: 0 writeConflicts: 0 numYields: 0 locks: {
    Global: {
    acquireCount: {
        r: 1
    }
    },
    MMAPV1Journal: {
    acquireCount: {
        r: 1
    }
    },
    Database: {
    acquireCount: {
        r: 1
    }
    },
    Collection: {
    acquireCount: {
        R: 1
    }
    }
}
1648 ms