CloudWatch Metrics

A cloudwatch metrics:

Represents a time-ordered set of data points published to CloudWatch.
It is similar to a variable to monitor, with data points as values of that variable over time.
data points can come from any application in AWS
AWS services send metrics to CloudWatch
Can send custom metrics to CloudWatch also
Can add data points in any order or at any rate
Retrieve statistics about data points as an ordered set of time-series data.
Metrics are specific to a Region of creation
Metrics cannot be deleted,
Automatically expire after 15 months if no new data published to them.
They expire on a rolling basis; as new data points come in, data older than 15 months drops.
Metrics are uniquely defined by a name, a namespace, and zero or more dimensions.
Each data point in a metric has a time stamp, and (optionally) a unit of measure.

Each metric data point must link with a time stamp.
The time stamp can be up to two weeks in the past
up to two hours into the future.
If no time stamp is given, CloudWatch creates a time stamp on time data point receive.
Time stamps are dateTime objects
Coordinated Universal Time (UTC) is recommended
When you retrieve statistics from CloudWatch, all times are in UTC
CloudWatch alarms check metrics based on the current time in UTC.

This retains metric data as follows:

Data points with a period of less than 60 seconds are available for 3 hours. Also called as high-resolution custom metrics.
Data points with a period of 60 seconds (1 minute) are available for 15 days
And, Data points with a period of 300 seconds (5 minute) are available for 63 days
Data points with a period of 3600 seconds (1 hour) are available for 455 days (15 months)

Each statistic has a unit of measure.
Example units – Bytes, Seconds, Count, and Percent.
Can specify a unit when you create a custom metric.
If not specified, CloudWatch uses None as the unit.
CloudWatch attaches no significance to a unit internally
Metric data points that specify a unit of measure aggregates separately.
Statistics without specifying a unit, CloudWatch aggregates all data points of the same unit together.

Period is the length of time associate with a specific Amazon CloudWatch statistic.
Periods defined in seconds, and valid values for period are 1, 5, 10, 30, or any multiple of 60.
For period of six minutes, use 360 as the period value.
Can adjust how the data aggregates by varying the length of the period.
Only custom metrics that you define with a storage resolution of 1 second support sub-minute periods.
To retrieve statistics, specify a period, start time, and end time.
The default values for the start time and end time get you the last hour’s worth of statistics.
For statistics aggregated over the entire hour, specify a period of 3600.
aggregated statistics are stamped with the time corresponding to the beginning of the period.
Periods are also important for CloudWatch alarms.

CloudWatch aggregates statistics as per specified period length
publish as many data points as needed with same or similar time stamps.
CloudWatch aggregates them as per specified period length.
CloudWatch does not aggregate data across Regions.
For large datasets, insert a pre-aggregated dataset called a statistic set.
With statistic sets, gives Min, Max, Sum, and SampleCount for a number of data points.
CloudWatch doesn’t differentiate the source of a metric.
Treated as single metric, even if metric is published with same namespace and dimensions from different sources

A percentile indicates the relative standing of a value in a dataset.
example, the 95th percentile means that 95 percent of the data is lower than this value and 5 percent of the data is higher than this value.
Used to isolate anomalies.