Amazon Web Services 文档中描述的 Amazon Web Services 服务或功能可能因区域而异。要查看适用于中国区域的差异，请参阅 中国的 Amazon Web Services 服务入门 (PDF)。

要获得与亚马逊 Timestream 类似的功能 LiveAnalytics，可以考虑适用于 InfluxDB 的亚马逊 Timestream。它为实时分析提供了简化的数据摄取和个位数毫秒的查询响应时间。点击此处了解更多。

本文属于机器翻译版本。若本译文内容与英语原文存在差异，则一律以英文原文为准。

使用时间序列函数的查询

示例数据集和查询

您可以使用 Timestream LiveAnalytics 来了解和提高您的服务和应用程序的性能和可用性。以下是一个示例表，以及在该表上运行的示例查询。

该表ec2_metrics存储遥测数据，例如 CPU 利用率和来自 EC2 实例的其他指标。您可以查看下表。

查找过去 2 小时内特定 EC2 主机的 CPU 平均利用率、p90、p95 和 p99：

SELECT region, az, hostname, BIN(time, 15s) AS binned_timestamp,
    ROUND(AVG(measure_value::double), 2) AS avg_cpu_utilization,
    ROUND(APPROX_PERCENTILE(measure_value::double, 0.9), 2) AS p90_cpu_utilization,
    ROUND(APPROX_PERCENTILE(measure_value::double, 0.95), 2) AS p95_cpu_utilization,
    ROUND(APPROX_PERCENTILE(measure_value::double, 0.99), 2) AS p99_cpu_utilization
FROM "sampleDB".DevOps
WHERE measure_name = 'cpu_utilization'
    AND hostname = 'host-Hovjv'
    AND time > ago(2h)
GROUP BY region, hostname, az, BIN(time, 15s)
ORDER BY binned_timestamp ASC

找出与过去 2 小时内整个队列的平均 CPU 利用率相比 CPU 利用率高出 10% 或以上的 EC2 主机：

WITH avg_fleet_utilization AS (
    SELECT COUNT(DISTINCT hostname) AS total_host_count, AVG(measure_value::double) AS fleet_avg_cpu_utilization
    FROM "sampleDB".DevOps
    WHERE measure_name = 'cpu_utilization'
        AND time > ago(2h)
), avg_per_host_cpu AS (
    SELECT region, az, hostname, AVG(measure_value::double) AS avg_cpu_utilization
    FROM "sampleDB".DevOps
    WHERE measure_name = 'cpu_utilization'
        AND time > ago(2h)
    GROUP BY region, az, hostname
)
SELECT region, az, hostname, avg_cpu_utilization, fleet_avg_cpu_utilization
FROM avg_fleet_utilization, avg_per_host_cpu
WHERE avg_cpu_utilization > 1.1 * fleet_avg_cpu_utilization
ORDER BY avg_cpu_utilization DESC

查找过去 2 小时内特定 EC2主机以 30 秒为间隔的平均 CPU 使用率：

SELECT BIN(time, 30s) AS binned_timestamp, ROUND(AVG(measure_value::double), 2) AS avg_cpu_utilization
FROM "sampleDB".DevOps
WHERE measure_name = 'cpu_utilization'
    AND hostname = 'host-Hovjv'
    AND time > ago(2h)
GROUP BY hostname, BIN(time, 30s)
ORDER BY binned_timestamp ASC

找出过去 2 小时内特定 EC2主机以 30 秒为间隔分箱的平均 CPU 利用率，使用线性插值填充缺失值：

WITH binned_timeseries AS (
    SELECT hostname, BIN(time, 30s) AS binned_timestamp, ROUND(AVG(measure_value::double), 2) AS avg_cpu_utilization
    FROM "sampleDB".DevOps
    WHERE measure_name = 'cpu_utilization'
        AND hostname = 'host-Hovjv'
        AND time > ago(2h)
    GROUP BY hostname, BIN(time, 30s)
), interpolated_timeseries AS (
    SELECT hostname,
        INTERPOLATE_LINEAR(
            CREATE_TIME_SERIES(binned_timestamp, avg_cpu_utilization),
                SEQUENCE(min(binned_timestamp), max(binned_timestamp), 15s)) AS interpolated_avg_cpu_utilization
    FROM binned_timeseries
    GROUP BY hostname
)
SELECT time, ROUND(value, 2) AS interpolated_cpu
FROM interpolated_timeseries
CROSS JOIN UNNEST(interpolated_avg_cpu_utilization)

找出过去 2 小时内特定 EC2主机以 30 秒为间隔分箱的平均 CPU 利用率，并根据上次执行的观测值使用插值填充缺失值：

WITH binned_timeseries AS (
    SELECT hostname, BIN(time, 30s) AS binned_timestamp, ROUND(AVG(measure_value::double), 2) AS avg_cpu_utilization
    FROM "sampleDB".DevOps
    WHERE measure_name = 'cpu_utilization'
        AND hostname = 'host-Hovjv'
        AND time > ago(2h)
    GROUP BY hostname, BIN(time, 30s)
), interpolated_timeseries AS (
    SELECT hostname,
        INTERPOLATE_LOCF(
            CREATE_TIME_SERIES(binned_timestamp, avg_cpu_utilization),
                SEQUENCE(min(binned_timestamp), max(binned_timestamp), 15s)) AS interpolated_avg_cpu_utilization
    FROM binned_timeseries
    GROUP BY hostname
)
SELECT time, ROUND(value, 2) AS interpolated_cpu
FROM interpolated_timeseries
CROSS JOIN UNNEST(interpolated_avg_cpu_utilization)