Skip to content

fix(collector): honor custom Postgres and PgBouncer ports in OTel metrics configs #4535

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
andrewlecuyer wants to merge 1 commit into
base: main
Choose a base branch
from
+120 −42
Open

fix(collector): honor custom Postgres and PgBouncer ports in OTel metrics configs #4535

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions internal/collector/pgbouncer.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -191,8 +191,8 @@ func EnablePgBouncerMetrics(ctx context.Context, inCluster *v1beta1.PostgresClus
config.Receivers[SqlQuery] = map[string]any{
"driver": "postgres",
"datasource": fmt.Sprintf(
`host=localhost dbname=pgbouncer port=5432 user=%s password=${env:PGPASSWORD}`,
sqlQueryUsername),
`host=localhost dbname=pgbouncer port=%d user=%s password=${env:PGPASSWORD}`,
*inCluster.Spec.Proxy.PGBouncer.Port, sqlQueryUsername),
"queries": slices.Clone(pgBouncerMetricsQueries),
}

Expand Down
36 changes: 36 additions & 0 deletions internal/collector/pgbouncer_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,9 @@ import (
"gotest.tools/v3/assert"

"github.com/crunchydata/postgres-operator/internal/feature"
"github.com/crunchydata/postgres-operator/internal/initialize"
"github.com/crunchydata/postgres-operator/internal/naming"
"github.com/crunchydata/postgres-operator/internal/testing/cmp"
"github.com/crunchydata/postgres-operator/internal/testing/require"
"github.com/crunchydata/postgres-operator/pkg/apis/postgres-operator.crunchydata.com/v1beta1"
)
Expand Down Expand Up @@ -337,6 +339,10 @@ func TestEnablePgBouncerMetrics(t *testing.T) {

config := NewConfig(nil)
cluster := new(v1beta1.PostgresCluster)
cluster.Spec.Proxy = &v1beta1.PostgresProxySpec{
PGBouncer: &v1beta1.PGBouncerPodSpec{},
}
cluster.Spec.Proxy.PGBouncer.Default() // Sets Port to 5432
require.UnmarshalInto(t, &cluster.Spec, `{
instrumentation: {}
}`)
Expand Down Expand Up @@ -513,6 +519,10 @@ service:
config := NewConfig(testInstrumentationSpec())

cluster := new(v1beta1.PostgresCluster)
cluster.Spec.Proxy = &v1beta1.PostgresProxySpec{
PGBouncer: &v1beta1.PGBouncerPodSpec{},
}
cluster.Spec.Proxy.PGBouncer.Default() // Sets Port to 5432
cluster.Spec.Instrumentation = testInstrumentationSpec()

EnablePgBouncerMetrics(ctx, cluster, config, "test_user")
Expand Down Expand Up @@ -683,4 +693,30 @@ service:
`)

})

t.Run("CustomPort", func(t *testing.T) {
gate := feature.NewGate()
assert.NilError(t, gate.SetFromMap(map[string]bool{
feature.OpenTelemetryMetrics: true,
}))
ctx := feature.NewContext(context.Background(), gate)

config := NewConfig(nil)
cluster := new(v1beta1.PostgresCluster)
cluster.Spec.Proxy = &v1beta1.PostgresProxySpec{
PGBouncer: &v1beta1.PGBouncerPodSpec{},
}
cluster.Spec.Proxy.PGBouncer.Default()
cluster.Spec.Proxy.PGBouncer.Port = initialize.Int32(6432) // Override default
require.UnmarshalInto(t, &cluster.Spec, `{
instrumentation: {}
}`)

EnablePgBouncerMetrics(ctx, cluster, config, "test_user")

// Verify datasource contains correct port
receiver := config.Receivers["sqlquery"].(map[string]any)
datasource := receiver["datasource"].(string)
assert.Assert(t, cmp.Contains(datasource, "port=6432"))
})
}
90 changes: 50 additions & 40 deletions internal/collector/postgres_metrics.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,6 +59,21 @@ type metric struct {
ValueType string `json:"value_type,omitempty"`
}

// updateServerAttribute updates the server static_attribute in all metrics
// to use the actual Postgres port from the cluster spec.
func updateServerAttribute(metrics []queryMetrics, port int32) {
serverValue := fmt.Sprintf("localhost:%d", port)
for i := range metrics {
for j := range metrics[i].Metrics {
if metrics[i].Metrics[j].StaticAttributes != nil {
if _, ok := metrics[i].Metrics[j].StaticAttributes["server"]; ok {
metrics[i].Metrics[j].StaticAttributes["server"] = serverValue
}
}
}
}
}

func EnablePostgresMetrics(ctx context.Context, inCluster *v1beta1.PostgresCluster, config *Config) {
if OpenTelemetryMetricsEnabled(ctx, inCluster) {
log := logging.FromContext(ctx)
Expand Down Expand Up @@ -96,54 +111,47 @@ func EnablePostgresMetrics(ctx context.Context, inCluster *v1beta1.PostgresClust
}
}

// Always parse JSON to update static_attributes.server with actual port
var fiveSecondMetricsArr []queryMetrics
if err := json.Unmarshal(fiveSecondMetricsClone, &fiveSecondMetricsArr); err != nil {
log.Error(err, "error parsing five second postgres metrics")
}
updateServerAttribute(fiveSecondMetricsArr, *inCluster.Spec.Port)

var fiveMinuteMetricsArr []queryMetrics
if err := json.Unmarshal(fiveMinuteMetricsClone, &fiveMinuteMetricsArr); err != nil {
log.Error(err, "error parsing five minute postgres metrics")
}
updateServerAttribute(fiveMinuteMetricsArr, *inCluster.Spec.Port)

var fiveMinutePerDBMetricsArr []queryMetrics
if err := json.Unmarshal(fiveMinutePerDBMetricsClone, &fiveMinutePerDBMetricsArr); err != nil {
log.Error(err, "error parsing per-db postgres metrics")
}
updateServerAttribute(fiveMinutePerDBMetricsArr, *inCluster.Spec.Port)

// Remove any queries that user has specified in the spec
if inCluster.Spec.Instrumentation != nil &&
inCluster.Spec.Instrumentation.Metrics != nil &&
inCluster.Spec.Instrumentation.Metrics.CustomQueries != nil &&
inCluster.Spec.Instrumentation.Metrics.CustomQueries.Remove != nil {

// Convert json to array of queryMetrics objects
var fiveSecondMetricsArr []queryMetrics
err := json.Unmarshal(fiveSecondMetricsClone, &fiveSecondMetricsArr)
if err != nil {
log.Error(err, "error compiling five second postgres metrics")
}

// Remove any specified metrics from the five second metrics
fiveSecondMetricsArr = removeMetricsFromQueries(
inCluster.Spec.Instrumentation.Metrics.CustomQueries.Remove, fiveSecondMetricsArr)

// Convert json to array of queryMetrics objects
var fiveMinuteMetricsArr []queryMetrics
err = json.Unmarshal(fiveMinuteMetricsClone, &fiveMinuteMetricsArr)
if err != nil {
log.Error(err, "error compiling five minute postgres metrics")
}

// Remove any specified metrics from the five minute metrics
fiveMinuteMetricsArr = removeMetricsFromQueries(
inCluster.Spec.Instrumentation.Metrics.CustomQueries.Remove, fiveMinuteMetricsArr)

// Convert json to array of queryMetrics objects
var fiveMinutePerDBMetricsArr []queryMetrics
err = json.Unmarshal(fiveMinutePerDBMetricsClone, &fiveMinutePerDBMetricsArr)
if err != nil {
log.Error(err, "error compiling per-db postgres metrics")
}

// Remove any specified metrics from the five minute per-db metrics
fiveMinutePerDBMetricsArr = removeMetricsFromQueries(
inCluster.Spec.Instrumentation.Metrics.CustomQueries.Remove, fiveMinutePerDBMetricsArr)

// Convert back to json data
// The error return value can be ignored as the errchkjson linter
// deems the []queryMetrics to be a safe argument:
// https://github.com/breml/errchkjson
fiveSecondMetricsClone, _ = json.Marshal(fiveSecondMetricsArr)
fiveMinuteMetricsClone, _ = json.Marshal(fiveMinuteMetricsArr)
fiveMinutePerDBMetricsClone, _ = json.Marshal(fiveMinutePerDBMetricsArr)
}

// Always marshal back to JSON
// The error return value can be ignored as the errchkjson linter
// deems the []queryMetrics to be a safe argument:
// https://github.com/breml/errchkjson
fiveSecondMetricsClone, _ = json.Marshal(fiveSecondMetricsArr)
fiveMinuteMetricsClone, _ = json.Marshal(fiveMinuteMetricsArr)
fiveMinutePerDBMetricsClone, _ = json.Marshal(fiveMinutePerDBMetricsArr)

// Add Prometheus exporter
config.Exporters[Prometheus] = map[string]any{
"endpoint": "0.0.0.0:" + strconv.Itoa(PrometheusPort),
Expand All @@ -152,8 +160,8 @@ func EnablePostgresMetrics(ctx context.Context, inCluster *v1beta1.PostgresClust
config.Receivers[FiveSecondSqlQuery] = map[string]any{
"driver": "postgres",
"datasource": fmt.Sprintf(
`host=localhost dbname=postgres port=5432 user=%s password=${env:PGPASSWORD}`,
MonitoringUser),
`host=localhost dbname=postgres port=%d user=%s password=${env:PGPASSWORD}`,
*inCluster.Spec.Port, MonitoringUser),
"collection_interval": "5s",
// Give Postgres time to finish setup.
"initial_delay": "15s",
Expand All @@ -163,8 +171,8 @@ func EnablePostgresMetrics(ctx context.Context, inCluster *v1beta1.PostgresClust
config.Receivers[FiveMinuteSqlQuery] = map[string]any{
"driver": "postgres",
"datasource": fmt.Sprintf(
`host=localhost dbname=postgres port=5432 user=%s password=${env:PGPASSWORD}`,
MonitoringUser),
`host=localhost dbname=postgres port=%d user=%s password=${env:PGPASSWORD}`,
*inCluster.Spec.Port, MonitoringUser),
"collection_interval": "300s",
// Give Postgres time to finish setup.
"initial_delay": "15s",
Expand Down Expand Up @@ -209,8 +217,9 @@ func EnablePostgresMetrics(ctx context.Context, inCluster *v1beta1.PostgresClust
config.Receivers[receiverName] = map[string]any{
"driver": "postgres",
"datasource": fmt.Sprintf(
`host=localhost dbname=%s port=5432 user=%s password=${env:PGPASSWORD}`,
`host=localhost dbname=%s port=%d user=%s password=${env:PGPASSWORD}`,
db,
*inCluster.Spec.Port,
MonitoringUser),
"collection_interval": querySet.CollectionInterval,
// Give Postgres time to finish setup.
Expand All @@ -234,8 +243,9 @@ func EnablePostgresMetrics(ctx context.Context, inCluster *v1beta1.PostgresClust
config.Receivers[receiverName] = map[string]any{
"driver": "postgres",
"datasource": fmt.Sprintf(
`host=localhost dbname=%s port=5432 user=%s password=${env:PGPASSWORD}`,
`host=localhost dbname=%s port=%d user=%s password=${env:PGPASSWORD}`,
db,
*inCluster.Spec.Port,
MonitoringUser),
"collection_interval": "5m",
// Give Postgres time to finish setup.
Expand Down
32 changes: 32 additions & 0 deletions internal/collector/postgres_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,11 +6,13 @@ package collector

import (
"context"
"encoding/json"
"testing"

"gotest.tools/v3/assert"

"github.com/crunchydata/postgres-operator/internal/feature"
"github.com/crunchydata/postgres-operator/internal/initialize"
"github.com/crunchydata/postgres-operator/internal/postgres"
"github.com/crunchydata/postgres-operator/internal/testing/cmp"
"github.com/crunchydata/postgres-operator/internal/testing/require"
Expand Down Expand Up @@ -903,6 +905,7 @@ func TestEnablePostgresMetrics(t *testing.T) {

cluster := new(v1beta1.PostgresCluster)
cluster.Spec.PostgresVersion = 99
cluster.Default() // Sets Port to 5432
require.UnmarshalInto(t, &cluster.Spec, `{
instrumentation: {}
}`)
Expand Down Expand Up @@ -967,6 +970,7 @@ service:

cluster := new(v1beta1.PostgresCluster)
cluster.Spec.PostgresVersion = 99
cluster.Default() // Sets Port to 5432
cluster.Spec.Instrumentation = testInstrumentationSpec()

config := NewConfig(cluster.Spec.Instrumentation)
Expand Down Expand Up @@ -1024,6 +1028,34 @@ service:
- sqlquery/300s
`)
})

t.Run("CustomPort", func(t *testing.T) {
gate := feature.NewGate()
assert.NilError(t, gate.SetFromMap(map[string]bool{
feature.OpenTelemetryMetrics: true,
}))
ctx := feature.NewContext(context.Background(), gate)

cluster := new(v1beta1.PostgresCluster)
cluster.Spec.PostgresVersion = 99
cluster.Default()
cluster.Spec.Port = initialize.Int32(5433) // Override default
require.UnmarshalInto(t, &cluster.Spec, `{
instrumentation: {}
}`)

config := NewConfig(nil)
EnablePostgresMetrics(ctx, cluster, config)

// Verify datasource contains correct port
receiver := config.Receivers["sqlquery/5s"].(map[string]any)
datasource := receiver["datasource"].(string)
assert.Assert(t, cmp.Contains(datasource, "port=5433"))

// Verify static_attributes.server contains correct port
queries := receiver["queries"].(json.RawMessage)
assert.Assert(t, cmp.Contains(string(queries), `"server":"localhost:5433"`))
})
}

func TestPostgresParameters(t *testing.T) {
Expand Down
Loading