Amazon RDS Performance Insights examples using AWS CLI

To describe dimension keys

This example requests the names of all wait events. The data is summarized by event name, and the aggregate values of those events over the specified time period.

Command:

aws pi describe-dimension-keys --service-type RDS --identifier db-LKCGOBK26374TPTDFXOIWVCPPM --start-time 1527026400 --end-time 1527080400 --metric db.load.avg --group-by '{"Group":"db.wait_event"}'

Output:

{
    "AlignedEndTime": 1.5270804E9,
    "AlignedStartTime": 1.5270264E9,
    "Keys": [
        {
            "Dimensions": {"db.wait_event.name": "wait/synch/mutex/innodb/aurora_lock_thread_slot_futex"},
            "Total": 0.05906906851195666
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/io/aurora_redo_log_flush"},
            "Total": 0.015824722186149193
        },
        {
            "Dimensions": {"db.wait_event.name": "CPU"},
            "Total": 0.008014396230265477
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/io/aurora_respond_to_client"},
            "Total": 0.0036361612526204477
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/io/table/sql/handler"},
            "Total": 0.0019108398419382965
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/synch/cond/mysys/my_thread_var::suspend"},
            "Total": 8.533847837782684E-4
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/io/file/csv/data"},
            "Total": 6.864181956477376E-4
        },
        {
            "Dimensions": {"db.wait_event.name": "Unknown"},
            "Total": 3.895887056379051E-4
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/synch/mutex/sql/FILE_AS_TABLE::LOCK_shim_lists"},
            "Total": 3.710368625122906E-5
        },
        {
            "Dimensions": {"db.wait_event.name": "wait/lock/table/sql/handler"},
            "Total": 0
        }
    ]
}

To get resource metrics

This example requests data points for the db.wait_event dimension group, and for the db.wait_event.name dimension within that group. In the response, the relevant data points are grouped by the requested dimension (db.wait_event.name).

Command:

aws pi get-resource-metrics --service-type RDS --identifier db-LKCGOBK26374TPTDFXOIWVCPPM --start-time 1527026400 --end-time 1527080400 --period-in-seconds 300 --metric db.load.avg --metric-queries file://metric-queries.json

The arguments for --metric-queries are stored in a JSON file, metric-queries.json. Here are the contents of that file:

[
    {
        "Metric": "db.load.avg",
        "GroupBy": {
            "Group":"db.wait_event"
        }
    }
]

Output:

{
    "AlignedEndTime": 1.5270804E9,
    "AlignedStartTime": 1.5270264E9,
    "Identifier": "db-LKCGOBK26374TPTDFXOIWVCPPM",
    "MetricList": [
        {
            "Key": {
                "Metric": "db.load.avg"
            },
            "DataPoints": [
                {
                    "Timestamp": 1527026700.0,
                    "Value": 1.3533333333333333
                },
                {
                    "Timestamp": 1527027000.0,
                    "Value": 0.88
                },
                <...remaining output omitted...>
            ]
        },
        {
            "Key": {
                "Metric": "db.load.avg",
                "Dimensions": {
                    "db.wait_event.name": "wait/synch/mutex/innodb/aurora_lock_thread_slot_futex"
                }
            },
            "DataPoints": [
                {
                    "Timestamp": 1527026700.0,
                    "Value": 0.8566666666666667
                },
                {
                    "Timestamp": 1527027000.0,
                    "Value": 0.8633333333333333
                },
                <...remaining output omitted...>
            ],
        },
            <...remaining output omitted...>
    ]
}