ZIO telemetry is purely-functional and type-safe. It provides clients for OpenTracing, OpenCensus and OpenTelemetry.
ZIO Telemetry consists of the following projects:
In monolithic architecture, everything is in one place, and we know when a request starts and then how it goes through the components and when it finishes. We can obviously see what is happening with our request and where is it going. But, in distributed systems like microservice architecture, we cannot find out the story of a request through various services easily. This is where distributed tracing comes into play.
ZIO Telemetry is a purely functional client which helps up propagate context between services in a distributed environment.
In order to use this library, we need to add the following line in our build.sbt file if we want to use OpenTelemetry client:
libraryDependencies += "dev.zio" %% "zio-telemetry" % "3.0.0-RC2"And for using OpenTracing client we should add the following line in our build.sbt file:
libraryDependencies += "dev.zio" %% "zio-opentracing" % "3.0.0-RC2"In this example, we create two services, ProxyServer and BackendServer. When we call ProxyServer, the BackendServer will be called.
Note that we are going to use OpenTracing client for this example.
Here is a simplified diagram of our services:
┌────────────────┐
│ │
┌─────►│ Jaeger Backend │◄────┐
│ │ │ │
Tracing Data │ └────────────────┘ │ Tracing Data
│ │
┌────────┴─────────┐ ┌─────────┴────────┐
│ │ │ │
User Request──►│ Proxy Server ├────────►| Backend Server │
│ │ │ │
└──────────────────┘ └──────────────────┘
First of all we should add following dependencies to our build.sbt file:
object Versions {
val http4s = "0.21.24"
val jaeger = "1.6.0"
val sttp = "2.2.9"
val opentracing = "0.33.0"
val opentelemetry = "1.4.1"
val opencensus = "0.28.3"
val zipkin = "2.16.3"
val zio = "1.0.9"
val zioInteropCats = "2.5.1.0"
}
lazy val openTracingExample = Seq(
"org.typelevel" %% "cats-core" % "2.6.1",
"io.circe" %% "circe-generic" % "0.14.1",
"org.http4s" %% "http4s-core" % Versions.http4s,
"org.http4s" %% "http4s-blaze-server" % Versions.http4s,
"org.http4s" %% "http4s-dsl" % Versions.http4s,
"org.http4s" %% "http4s-circe" % Versions.http4s,
"io.jaegertracing" % "jaeger-core" % Versions.jaeger,
"io.jaegertracing" % "jaeger-client" % Versions.jaeger,
"io.jaegertracing" % "jaeger-zipkin" % Versions.jaeger,
"com.github.pureconfig" %% "pureconfig" % "0.16.0",
"com.softwaremill.sttp.client" %% "async-http-client-backend-zio" % Versions.sttp,
"com.softwaremill.sttp.client" %% "circe" % Versions.sttp,
"dev.zio" %% "zio-interop-cats" % Versions.zioInteropCats,
"io.zipkin.reporter2" % "zipkin-reporter" % Versions.zipkin,
"io.zipkin.reporter2" % "zipkin-sender-okhttp3" % Versions.zipkin
)Let's create a ZLayer for OpenTracing which provides us Jaeger tracer. Each microservice uses this dependency to send its tracing data to the Jaeger Backend:
import io.jaegertracing.Configuration
import io.jaegertracing.internal.samplers.ConstSampler
import io.jaegertracing.zipkin.ZipkinV2Reporter
import org.apache.http.client.utils.URIBuilder
import zio.ZLayer
import zio.clock.Clock
import zio.telemetry.opentracing.OpenTracing
import zipkin2.reporter.AsyncReporter
import zipkin2.reporter.okhttp3.OkHttpSender
object JaegerTracer {
def makeJaegerTracer(host: String, serviceName: String): ZLayer[Clock, Throwable, Clock with OpenTracing] =
OpenTracing.live(new Configuration(serviceName)
.getTracerBuilder
.withSampler(new ConstSampler(true))
.withReporter(
new ZipkinV2Reporter(
AsyncReporter.create(
OkHttpSender.newBuilder
.compressionEnabled(true)
.endpoint(
new URIBuilder()
.setScheme("http")
.setHost(host)
.setPath("/api/v2/spans")
.build.toString
)
.build
)
)
)
.build
) ++ Clock.live
}The BackendServer:
import io.opentracing.propagation.Format.Builtin.{HTTP_HEADERS => HttpHeadersFormat}
import io.opentracing.propagation.TextMapAdapter
import org.http4s._
import org.http4s.dsl.Http4sDsl
import org.http4s.server.Router
import org.http4s.server.blaze.BlazeServerBuilder
import org.http4s.syntax.kleisli._
import zio.clock.Clock
import zio.interop.catz._
import zio.telemetry.opentracing._
import JaegerTracer.makeJaegerTracer
import zio.{ExitCode, ZEnv, ZIO}
import scala.jdk.CollectionConverters._
object BackendServer extends CatsApp {
type AppTask[A] = ZIO[Clock, Throwable, A]
val dsl: Http4sDsl[AppTask] = Http4sDsl[AppTask]
import dsl._
override def run(args: List[String]): ZIO[ZEnv, Nothing, ExitCode] =
ZIO.runtime[Clock].flatMap { implicit runtime =>
BlazeServerBuilder[AppTask](runtime.platform.executor.asEC)
.bindHttp(port = 9000, host = "0.0.0.0")
.withHttpApp(
Router[AppTask](mappings = "/" ->
HttpRoutes.of[AppTask] { case request@GET -> Root =>
ZIO.unit
.spanFrom(
format = HttpHeadersFormat,
carrier = new TextMapAdapter(request.headers.toList.map(h => h.name.value -> h.value).toMap.asJava),
operation = "GET /"
)
.provide(makeJaegerTracer(host = "0.0.0.0:9411", serviceName = "backend-service")) *> Ok("Ok!")
}
).orNotFound
)
.serve
.compile
.drain
}.exitCode
}And the ProxyServer which calls the BackendServer:
import cats.effect.{ExitCode => catsExitCode}
import io.opentracing.propagation.Format.Builtin.{HTTP_HEADERS => HttpHeadersFormat}
import io.opentracing.propagation.TextMapAdapter
import io.opentracing.tag.Tags
import org.http4s.HttpRoutes
import org.http4s.dsl.Http4sDsl
import org.http4s.server.Router
import org.http4s.server.blaze.BlazeServerBuilder
import org.http4s.syntax.kleisli._
import sttp.client.asynchttpclient.zio.AsyncHttpClientZioBackend
import sttp.client.basicRequest
import sttp.model.Uri
import zio.clock.Clock
import zio.interop.catz._
import zio.telemetry.opentracing.OpenTracing
import JaegerTracer.makeJaegerTracer
import zio.{ExitCode, UIO, ZEnv, ZIO}
import scala.collection.mutable
import scala.jdk.CollectionConverters._
object ProxyServer extends CatsApp {
type AppTask[A] = ZIO[Clock, Throwable, A]
private val backend = AsyncHttpClientZioBackend()
override def run(args: List[String]): ZIO[ZEnv, Nothing, ExitCode] =
ZIO.runtime[Clock].flatMap { implicit runtime =>
implicit val ec = runtime.platform.executor.asEC
BlazeServerBuilder[AppTask](ec)
.bindHttp(port = 8080, host = "0.0.0.0")
.withHttpApp(
Router[AppTask](mappings = "/" -> {
val dsl: Http4sDsl[AppTask] = Http4sDsl[AppTask]
import dsl._
HttpRoutes.of[AppTask] { case GET -> Root =>
(for {
_ <- OpenTracing.tag(Tags.SPAN_KIND.getKey, Tags.SPAN_KIND_CLIENT)
_ <- OpenTracing.tag(Tags.HTTP_METHOD.getKey, GET.name)
_ <- OpenTracing.setBaggageItem("proxy-baggage-item-key", "proxy-baggage-item-value")
buffer = new TextMapAdapter(mutable.Map.empty[String, String].asJava)
_ <- OpenTracing.provide(HttpHeadersFormat, buffer)
headers <- extractHeaders(buffer)
res <-
backend.flatMap { implicit backend =>
basicRequest.get(Uri("0.0.0.0", 9000).path("/")).headers(headers).send()
}.map(_.body)
.flatMap {
case Right(_) => Ok("Ok!")
case Left(_) => Ok("Oops!")
}
} yield res)
.root(operation = "GET /")
.provide(
makeJaegerTracer(host = "0.0.0.0:9411", serviceName = "proxy-server")
)
}
}).orNotFound
)
.serve
.compile[AppTask, AppTask, catsExitCode]
.drain
.as(ExitCode.success)
}.exitCode
private def extractHeaders(adapter: TextMapAdapter): UIO[Map[String, String]] = {
val m = mutable.Map.empty[String, String]
UIO(adapter.forEach { entry =>
m.put(entry.getKey, entry.getValue)
()
}).as(m.toMap)
}
}First, we run the following command to start Jaeger backend:
docker run -d --name jaeger \
-e COLLECTOR_ZIPKIN_HTTP_PORT=9411 \
-p 5775:5775/udp \
-p 6831:6831/udp \
-p 6832:6832/udp \
-p 5778:5778 \
-p 16686:16686 \
-p 14268:14268 \
-p 9411:9411 \
jaegertracing/all-in-one:1.6It's time to run Backend and Proxy servers. After starting these two, we can start calling ProxyServer:
curl -X GET http://0.0.0.0:8080/Now we can check the Jaeger service (http://localhost:16686/) to see the result.
- Trace your microservices with ZIO by Marek Kadek (September 2021)
Learn more on the ZIO Telemetry homepage!
For the general guidelines, see ZIO contributor's guide.
See the Code of Conduct
Come chat with us on 