Provide a backend for the metrics
facade
crate, pushing metrics to CloudWatch.
Credentials for AWS needs to be available in the environment, see AWS docs on setting up AWS credentials
cargo add -s metrics metrics_cloudwatch
#[tokio::main]
async fn main() {
let sdk_config = aws_config::defaults(aws_config::BehaviorVersion::latest())
.load()
.await;
let client = aws_sdk_cloudwatch::Client::new(&sdk_config);
// Initialize the backend
tokio::spawn(metrics_cloudwatch::Builder::new()
.cloudwatch_namespace("my-namespace")
.init_future(client, metrics::set_global_recorder));
metrics::counter!("requests").increment(1);
}
Any labels specified will map to Cloudwatch dimensions
metrics::histogram!("histogram", "dimension_name" => "dimension_value").record(100.0);
Specifying the empty string for the value will remove the default dimension of the same name from the metric.
metrics::histogram!("histogram", "dimension_name" => "").record(100.0);
The special @unit
label accepts a metrics_cloudwatch::Unit
which specifies the unit for the metric (the unit can also be specified when registering the metric). Other @
prefixed labels are ignored.
metrics::histogram!("histogram", "@unit" => metrics_cloudwatch::Unit::Seconds).record(100.0);
The CloudWatch metrics API imposes some limitations.
- Max 30 labels (dimensions) per metric
- Max 150 unique histogram values (used by
timing!()
andvalue!()
) per API call. Going beyond this works but will incur one API call per batch of 150 unique values. Could be a good idea to measure timing in milliseconds rather than nanoseconds, to keep down the number of unique values.