Gin
Send distributed traces and metrics to Last9 from Golang Gin application using OpenTelemetry
Introduction
Gin is a web framework written in Go (Golang). This comprehensive guide will help you instrument your Gin application with OpenTelemetry and smoothly send the traces to Last9. You can also check out the example application on GitHub↗.
Pre-requisites
- You have a Gin application.
- You have signed up for Last9, and obtained the following OTLP credentials from the Integrations page:
endpoint
auth_header
Install OpenTelemetry packages
To install the required packages, run the following command:
go get -u go.opentelemetry.io/otel
go get -u go.opentelemetry.io/otel/sdk
go get -u go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc
go get -u go.opentelemetry.io/contrib/instrumentation/github.com/gin-gonic/gin/otelgin
go get -u go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc
go get -u go.opentelemetry.io/otel/sdk/metric
go get -u go.nhat.io/otelsql
For setting up Redis instrumentation, first verify which go-redis
version you are using.
If you are using go-redis
v8 then
go get -u github.com/go-redis/redis/extra/redisotel/v8
If you are using go-redis
v9 then
go get github.com/redis/go-redis/extra/redisotel/v9
Using these packages, you can instrument your Gin application to send traces and metrics to Last9.
Traces
This application generates traces for the following:
For HTTP requests, wrap the Gin router with the otelgin.Middleware
middleware.
Refer to main.go for more details.
For database queries, use the otelsql
package to wrap the sql.DB
object.
Refer to
initDB()
in users/controller.go for more details.
For Redis commands, use the redisotel
package to wrap the redis.Client
object.
Refer to
initRedis()
in users/controller.go for more details.
Metrics
The application also generates metrics for the following:
- Database queries using otelsql
Set the environment variables
Set the following environment variables:
export OTEL_SERVICE_NAME=gin-app-service
export OTEL_EXPORTER_OTLP_ENDPOINT=<endpoint>
export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Basic <auth_header>"
These environment variables are used to configure the OpenTelemetry SDK to send traces and metrics to Last9.
Instrument HTTP requests
In your Gin application, create a new file instrumentation.go
and add the following code:
Refer to main.go for more details.
package main
import (
"context"
"go.opentelemetry.io/otel"
"go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc"
"go.opentelemetry.io/otel/propagation"
sdktrace "go.opentelemetry.io/otel/sdk/trace"
semconv "go.opentelemetry.io/otel/semconv/v1.26.0"
"go.opentelemetry.io/otel/trace"
)
type Instrumentation struct {
TracerProvider *sdktrace.TracerProvider
Tracer trace.Tracer
}
func initTracerProvider() *sdktrace.TracerProvider {
exporter, err := otlptracegrpc.New(context.Background())
if err != nil {
panic(err)
}
attr := resource.WithAttributes(
semconv.DeploymentEnvironmentKey.String("production"), // You can change this value to "development" or "staging" or you can get the value from the environment variables
semconv.ServiceNameKey.String("gin-server"), // You can change this value to the name of your service
// You can add more attributes here
)
resources, err := resource.New(context.Background(),
resource.WithFromEnv(),
resource.WithTelemetrySDK(),
resource.WithProcess(),
resource.WithOS(),
resource.WithContainer(),
resource.WithHost(),
attr)
if err != nil {
panic(err)
}
tp := sdktrace.NewTracerProvider(
sdktrace.WithBatcher(exporter),
sdktrace.WithResource(resources),
)
otel.SetTracerProvider(tp)
otel.SetTextMapPropagator(propagation.NewCompositeTextMapPropagator(propagation.TraceContext{}, propagation.Baggage{}))
return tp
}
func NewInstrumentation() *Instrumentation {
tp := initTracerProvider()
return &Instrumentation{
TracerProvider: tp,
Tracer: tp.Tracer("gin-server"),
}
}
The above code configures the OpenTelemetry SDK to use the OTLP exporter and initializes the TracerProvider.
Next, at the entry point of your application, add the following code to instrument your application: Refer to main.go for more details.
i := NewInstrumentation()
defer func() {
if err := i.TracerProvider.Shutdown(context.Background()); err != nil {
log.Printf("Error shutting down tracer provider: %v", err)
}
}()
Now, add the otel gin middleware to your application:
import (
// other imports
"go.opentelemetry.io/contrib/instrumentation/github.com/gin-gonic/gin/otelgin"
)
// other code
r := gin.Default()
// Make sure this is added before any routes are defined
r.Use(otelgin.Middleware("gin-server"))
// other code
Refer to main.go for more details.
Use the tracer
object to create spans in your application as follows with custom attributes such as user id:
Refer to users/controller.go for more details.
_, span := u.tracer.Start(c.Request.Context(), "UpdateUser", oteltrace.WithAttributes(
attribute.String("user.id", c.Param("id")),
))
defer span.End()
Instrument database operations
Instrumenting with sql.DB
Add the following code to instrument the database queries.
It uses the otelsql
package to wrap the sql.DB
object and emit traces and metrics for database queries and connections.
Refer to users/controller.go for more details.
func initDB() (*sql.DB, error) {
driverName, err := otelsql.Register("postgres",
// Read more about the options here: https://github.com/nhatthm/otelsql?tab=readme-ov-file#options
otelsql.AllowRoot(),
otelsql.TraceQueryWithoutArgs(),
otelsql.TraceRowsClose(),
otelsql.TraceRowsAffected(),
otelsql.WithDatabaseName("otel_demo"), // database name
otelsql.WithSystem(semconv.DBSystemPostgreSQL),
)
if err != nil {
return nil, fmt.Errorf("failed to register driver: %v", err)
}
db, err := sql.Open(driverName, dsnName)
if err != nil {
return nil, fmt.Errorf("failed to connect to database: %v", err)
}
// Record stats to expose metrics
if err := otelsql.RecordStats(db); err != nil {
return nil, err
}
return db, nil
}
Instrumenting with pgx
For database instrumentation where pgx is used, we use otelpgx to wrap the pgx connection pool. Add the following code to instrument the database queries.
var connString = os.Getenv("DATABASE_URL")
cfg, err := pgxpool.ParseConfig(connString)
if err != nil {
fmt.Fprintf(os.Stderr, "create connection pool: %w", err)
os.Exit(1)
}
// Add the tracer to the connection pool configuration
cfg.ConnConfig.Tracer = otelpgx.NewTracer()
// Create a new connection pool with the tracer
conn, err = pgxpool.NewWithConfig(context.Background(), cfg)
if err != nil {
fmt.Fprintf(os.Stderr, "Unable to connection to database: %v\n", err)
os.Exit(1)
}
Refer to the complete example using pgx adapter and Otel instrumentation here.
Instrument Redis operations
Note that between redis-go
versions v8 and v9 the import paths are changed.
Add the following code to instrument the Redis operations. It uses the redisotel
package to wrap the redis.Client
object and emit traces for Redis commands.
go-redis v9
If you are using go-redis
v9 then use the following code.
Refer to users/controller.go for more details.
func initRedis() *redis.Client {
rdb := redis.NewClient(&redis.Options{
Addr: "localhost:6379", // Update this with your Redis server address
})
// Setup traces for redis instrumentation
if err := redisotel.InstrumentTracing(rdb); err != nil {
log.Fatalf("failed to instrument traces for Redis client: %v", err)
return nil
}
return rdb
}
go-redis v8
If you are using go-redis
v8 then use the following code.
func initRedis() *redis.Client {
rdb := redis.NewClient(&redis.Options{
Addr: "localhost:6379", // Update this with your Redis server address
})
// Setup traces for redis instrumentation
rdb.AddHook(redisotel.NewTracingHook())
return rdb
}
Refer to main.go for more details.
Run the application
Start your Gin application by running the following command:
go run main.go
This will start the Gin application and the OpenTelemetry SDK will automatically collect traces and metrics from the Gin application. These traces and metrics will be sent to Last9 automatically based on the environment variables set.
View traces and metrics in Last9
After running the Gin app, you can visualize the traces and metrics in Last9's APM dashboard.
Troubleshooting
Please get in touch with us on Discord or Email if you have any questions.