This is a proof-of-concept tracing layer, that propagates tracing spans to AWS X-Ray.
First, add tracing, tracing_subscriber, tracing-xray crates as dependencies:
tracing = "0.1.13"
tracing-subscriber = "0.3.11"
tracing-xray = { git = "https://github.com/jswrenn/tracing-xray.git" }Then, add tracing_xray::Layer as a tracing layer:
#[tokio::main]
pub async fn main() -> mini_redis::Result<()> {
use tracing_subscriber::{registry::Registry, prelude::*};
let _subscriber = Registry::default()
.with(tracing_xray::Layer::new("YOUR-SERVICE-NAME".to_owned()).await?)
.init();
...
}Finally, use the #[tracing::instrument] macro to trace key functions in your application. For a complete example, see here.
X-Ray segments are emitted for all tracing spans with an explicit aws.xray.trace_id field; e.g.:
#[instrument(
skip(db),
fields(
aws.xray.trace_id = &tracing_xray::trace_id::new().as_str(),
),
)]
async fn handle(db: &Db, request: Request) -> crate::Result<()> {
...
}X-Ray subsegments are emitted for all tracing spans that have an ancestor with an explicit aws.xray.trace_id field.
The fields of tracing spans are translated into X-Ray annotations and metadata. To emit an X-Ray annotation, prefix your field's name with aws.xray.annotations.. All other fields (except for aws.xray.trace_id) are emitted as metadata.
To test whether X-Ray (sub)segments are being emitted, use netcat to listen to UDP traffic on local port 2000 while your application is doing work; e.g.:
$ nc -ul 2000
{"format": "json", "version": 1}
{"name":"mini-redis-server","id":"0000000000000006","start_time":1650387854.3217046,"trace_id":"1-625eeb8e-1ea68f58158546959418a682","metadata":{"tracing.file":"src/server.rs","tracing.line":329},"annotations":{"tracing.name":"run","tracing.target":"mini_redis::server"},"in_progress":true}
{"format": "json", "version": 1}
{"name":"apply","id":"000000c000000007","start_time":1650387854.3320353,"trace_id":"1-625eeb8e-45b511399406edc3a114d091","parent_id":"0000004000000001","type":"subsegment","metadata":{"tracing.file":"src/cmd/set.rs","tracing.line":127},"annotations":{"tracing.target":"mini_redis::cmd::set"},"end_time":1650387854.3321207}
{"format": "json", "version": 1}
{"name":"apply","id":"000000c000000007","start_time":1650387854.3320353,"trace_id":"1-625eeb8e-45b511399406edc3a114d091","parent_id":"0000004000000001","type":"subsegment","metadata":{"tracing.file":"src/cmd/set.rs","tracing.line":127},"annotations":{"tracing.target":"mini_redis::cmd::set"},"in_progress":true}
{"format": "json", "version": 1}
{"name":"mini-redis-server","id":"0000004000000001","start_time":1650387854.3205793,"trace_id":"1-625eeb8e-45b511399406edc3a114d091","metadata":{"tracing.file":"src/server.rs","tracing.line":329},"annotations":{"tracing.name":"run","tracing.target":"mini_redis::server"},"end_time":1650387854.332266}
This assumes you have an AWS account already created.
First, in the AWS console, create a new IAM user, and give it the Permissions policy called AWSXRayDaemonWriteAccess (this policy already exists, you don't have to create it).
Then, grab the Access Key ID and Secret Access Key for your user. You will want to create a credentials file called ``~/.aws/credentials`:
[default]
aws_access_key_id = <id>
aws_secret_access_key = <key>
Next, download the X-Ray daemon.
Finally, run the X-Ray daemon with ./xray. Use -n to set the region, i.e, -n us-east-2). In the terminal running the X-Ray daemon, you should see something like:
[Info] Successfully sent batch of 50 segments (0.024 seconds)
In the AWS console, you can go to the X-Ray web interface to see your traces (make sure you are in the correct region).
This crate is a proof-of-concept. It is not supported, nor production-ready. Its implementation is deliberately simple.