Add Azure platform detection to the translator

cfg/envconfig/envconfig.go

@@ -26,6 +26,7 @@ const (
 	RunInContainer              = "RUN_IN_CONTAINER"
 	RunAsHostProcessContainer   = "RUN_AS_HOST_PROCESS_CONTAINER"
 	RunInAWS                    = "RUN_IN_AWS"
+	RunInAKS                    = "RUN_IN_AKS"
 	RunWithIRSA                 = "RUN_WITH_IRSA"
 	RunWithSELinux              = "RUN_WITH_SELINUX"
 	RunInROSA                   = "RUN_IN_ROSA"
@@ -89,6 +90,11 @@ func IsRunningInROSA() bool {
 	return os.Getenv(RunInROSA) == TrueValue
 }
 
+// IsRunningInAKS reports whether RUN_IN_AKS is set (by the AKS Helm chart) so AKS is detected without a metadata probe.
+func IsRunningInAKS() bool {
+	return os.Getenv(RunInAKS) == TrueValue
+}
+
 func GetLogsBackpressureMode() string {
 	return os.Getenv(CWAgentLogsBackpressureMode)
 }

service/defaultcomponents/components.go

@@ -17,6 +17,7 @@ import (
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/headerssetterextension"
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/healthcheckextension"
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/observer/ecsobserver"
+	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/oidctokenextension"
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/pprofextension"
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/sigv4authextension"
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/storage/filestorage"
@@ -187,6 +188,7 @@ func Factories() (otelcol.Factories, error) {
 		ecsobserver.NewFactory(),
 		filestorage.NewFactory(),
 		healthcheckextension.NewFactory(),
+		oidctokenextension.NewFactory(),
 		pprofextension.NewFactory(),
 		sigv4authextension.NewFactory(),
 		zpagesextension.NewFactory(),

service/defaultcomponents/components_test.go

@@ -126,6 +126,7 @@ func TestComponents(t *testing.T) {
 		"health_check",
 		"k8smetadata",
 		"nodemetadatacache",
+		"oidctoken",
 		"pprof",
 		"server",
 		"sigv4auth",

tool/paths/paths.go

@@ -11,6 +11,7 @@ const (
 	ENV            = "env-config.json"
 	AGENT_LOG_FILE = "amazon-cloudwatch-agent.log"
 	JMXJarName     = "opentelemetry-jmx-metrics.jar"
+	OIDCToken      = ".oidc-token"
 )
 
 var (
@@ -24,4 +25,5 @@ var (
 	TranslatorBinaryPath string
 	AgentBinaryPath      string
 	JMXJarPath           string
+	OIDCTokenPath        string
 )

tool/paths/paths_unix.go

@@ -31,4 +31,5 @@ func init() {
 	TranslatorBinaryPath = filepath.Join(AgentDir, "bin", TranslatorBinaryName)
 	AgentBinaryPath = filepath.Join(AgentDir, "bin", AgentBinaryName)
 	JMXJarPath = filepath.Join(AgentDir, "bin", JMXJarName)
+	OIDCTokenPath = filepath.Join(AgentDir, "etc", OIDCToken)
 }

tool/paths/paths_windows.go

@@ -53,4 +53,5 @@ func init() {
 	TranslatorBinaryPath = filepath.Join(AgentRootDir, TranslatorBinaryName)
 	AgentBinaryPath = filepath.Join(AgentRootDir, AgentBinaryName)
 	JMXJarPath = filepath.Join(AgentRootDir, JMXJarName)
+	OIDCTokenPath = filepath.Join(AgentConfigDir, OIDCToken)
 }

translator/config/mode.go

@@ -20,11 +20,19 @@ const (
 	ModeK8sOnPrem = "K8sOnPrem"
 )
 
+// Azure platform modes: ModeAzureVM is host-level (like ModeEC2), ModeAKS is Kubernetes-level (like ModeEKS).
+const (
+	ModeAzureVM = "AzureVM"
+	ModeAKS     = "AKS"
+)
+
 const (
 	ShortModeEC2       = "EC2"
 	ShortModeOnPrem    = "OP"
 	ShortModeWithIRSA  = "WI"
 	ShortModeEKS       = "EKS"
 	ShortModeK8sEC2    = "K8E"
 	ShortModeK8sOnPrem = "K8OP"
+	ShortModeAzureVM   = "AZVM"
+	ShortModeAKS       = "AKS"
 )

translator/context/context.go

@@ -129,8 +129,11 @@ func (ctx *Context) SetMode(mode string) {
 	case config.ModeWithIRSA:
 		ctx.mode = config.ModeWithIRSA
 		ctx.shortMode = config.ShortModeWithIRSA
+	case config.ModeAzureVM:
+		ctx.mode = config.ModeAzureVM
+		ctx.shortMode = config.ShortModeAzureVM
 	default:
-		log.Panicf("Invalid mode %s. Valid mode values are %s, %s, %s, and %s.", mode, config.ModeEC2, config.ModeOnPrem, config.ModeOnPremise, config.ModeWithIRSA)
+		log.Panicf("Invalid mode %s. Valid mode values are %s, %s, %s, %s, and %s.", mode, config.ModeEC2, config.ModeOnPrem, config.ModeOnPremise, config.ModeWithIRSA, config.ModeAzureVM)
 	}
 }
 
@@ -145,6 +148,9 @@ func (ctx *Context) SetKubernetesMode(mode string) {
 	case config.ModeK8sOnPrem:
 		ctx.kubernetesMode = config.ModeK8sOnPrem
 		ctx.shortMode = config.ShortModeK8sOnPrem
+	case config.ModeAKS:
+		ctx.kubernetesMode = config.ModeAKS
+		ctx.shortMode = config.ShortModeAKS
 	default:
 		ctx.kubernetesMode = ""
 	}

translator/context/context_test.go

@@ -0,0 +1,91 @@
+// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+// SPDX-License-Identifier: MIT
+
+package context
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+
+	"github.com/aws/amazon-cloudwatch-agent/translator/config"
+)
+
+func TestSetMode_Azure(t *testing.T) {
+	ResetContext()
+	ctx := CurrentContext()
+
+	ctx.SetMode(config.ModeAzureVM)
+	assert.Equal(t, config.ModeAzureVM, ctx.Mode())
+	assert.Equal(t, config.ShortModeAzureVM, ctx.ShortMode())
+}
+
+func TestSetMode_ExistingModesUnchanged(t *testing.T) {
+	cases := map[string]struct {
+		mode      string
+		wantShort string
+	}{
+		"EC2":      {config.ModeEC2, config.ShortModeEC2},
+		"OnPrem":   {config.ModeOnPrem, config.ShortModeOnPrem},
+		"WithIRSA": {config.ModeWithIRSA, config.ShortModeWithIRSA},
+	}
+	for name, tc := range cases {
+		t.Run(name, func(t *testing.T) {
+			ResetContext()
+			ctx := CurrentContext()
+			ctx.SetMode(tc.mode)
+			assert.Equal(t, tc.mode, ctx.Mode())
+			assert.Equal(t, tc.wantShort, ctx.ShortMode())
+		})
+	}
+}
+
+func TestSetKubernetesMode_AKS(t *testing.T) {
+	ResetContext()
+	ctx := CurrentContext()
+
+	ctx.SetKubernetesMode(config.ModeAKS)
+	assert.Equal(t, config.ModeAKS, ctx.KubernetesMode())
+	assert.Equal(t, config.ShortModeAKS, ctx.ShortMode())
+}
+
+func TestSetKubernetesMode_ExistingModesUnchanged(t *testing.T) {
+	cases := map[string]struct {
+		mode      string
+		wantShort string
+	}{
+		"EKS":       {config.ModeEKS, config.ShortModeEKS},
+		"K8sEC2":    {config.ModeK8sEC2, config.ShortModeK8sEC2},
+		"K8sOnPrem": {config.ModeK8sOnPrem, config.ShortModeK8sOnPrem},
+	}
+	for name, tc := range cases {
+		t.Run(name, func(t *testing.T) {
+			ResetContext()
+			ctx := CurrentContext()
+			ctx.SetKubernetesMode(tc.mode)
+			assert.Equal(t, tc.mode, ctx.KubernetesMode())
+			assert.Equal(t, tc.wantShort, ctx.ShortMode())
+		})
+	}
+}
+
+func TestSetKubernetesMode_UnknownClears(t *testing.T) {
+	ResetContext()
+	ctx := CurrentContext()
+	// A previously-set valid kubernetes mode is cleared by an unknown one.
+	ctx.SetKubernetesMode(config.ModeAKS)
+	ctx.SetKubernetesMode("not-a-real-mode")
+	assert.Equal(t, "", ctx.KubernetesMode())
+	// The default branch intentionally keeps shortMode (SetMode sets it first); with no host mode set it retains the last k8s shortMode.
+	assert.Equal(t, config.ShortModeAKS, ctx.ShortMode())
+}
+
+func TestSetKubernetesMode_EmptyPreservesHostShortMode(t *testing.T) {
+	ResetContext()
+	ctx := CurrentContext()
+	// Real startup ordering: host mode first, then empty (non-Kubernetes) mode; the host shortMode must survive.
+	ctx.SetMode(config.ModeAzureVM)
+	ctx.SetKubernetesMode("")
+	assert.Equal(t, "", ctx.KubernetesMode())
+	assert.Equal(t, config.ShortModeAzureVM, ctx.ShortMode())
+}

translator/tocwconfig/sampleConfig/opentelemetry/combined_v1_v2_ec2_config.yaml

@@ -926,23 +926,23 @@ processors:
               error_mode: ignore
               statements:
                 - set(resource.attributes["service.name"], "unknown_service") where resource.attributes["service.name"] == nil
-                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil
+                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil and resource.attributes["cloud.platform"] == "aws_ec2"
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], resource.attributes["ec2.tag.aws:autoscaling:groupName"]], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["ec2.tag.aws:autoscaling:groupName"] != nil
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], "default"], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["cloud.platform"] != nil
         metric_statements:
             - context: resource
               error_mode: ignore
               statements:
                 - set(resource.attributes["service.name"], "unknown_service") where resource.attributes["service.name"] == nil
-                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil
+                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil and resource.attributes["cloud.platform"] == "aws_ec2"
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], resource.attributes["ec2.tag.aws:autoscaling:groupName"]], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["ec2.tag.aws:autoscaling:groupName"] != nil
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], "default"], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["cloud.platform"] != nil
         trace_statements:
             - context: resource
               error_mode: ignore
               statements:
                 - set(resource.attributes["service.name"], "unknown_service") where resource.attributes["service.name"] == nil
-                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil
+                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil and resource.attributes["cloud.platform"] == "aws_ec2"
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], resource.attributes["ec2.tag.aws:autoscaling:groupName"]], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["ec2.tag.aws:autoscaling:groupName"] != nil
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], "default"], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["cloud.platform"] != nil
     transform/logs_cleanup:

translator/tocwconfig/sampleConfig/opentelemetry/dbi_config_linux.yaml

@@ -729,23 +729,23 @@ processors:
               error_mode: ignore
               statements:
                 - set(resource.attributes["service.name"], "unknown_service") where resource.attributes["service.name"] == nil
-                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil
+                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil and resource.attributes["cloud.platform"] == "aws_ec2"
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], resource.attributes["ec2.tag.aws:autoscaling:groupName"]], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["ec2.tag.aws:autoscaling:groupName"] != nil
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], "default"], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["cloud.platform"] != nil
         metric_statements:
             - context: resource
               error_mode: ignore
               statements:
                 - set(resource.attributes["service.name"], "unknown_service") where resource.attributes["service.name"] == nil
-                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil
+                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil and resource.attributes["cloud.platform"] == "aws_ec2"
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], resource.attributes["ec2.tag.aws:autoscaling:groupName"]], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["ec2.tag.aws:autoscaling:groupName"] != nil
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], "default"], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["cloud.platform"] != nil
         trace_statements:
             - context: resource
               error_mode: ignore
               statements:
                 - set(resource.attributes["service.name"], "unknown_service") where resource.attributes["service.name"] == nil
-                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil
+                - set(resource.attributes["cloud.resource_id"], Concat(["arn:aws:ec2", resource.attributes["cloud.region"], resource.attributes["cloud.account.id"], Concat(["instance", resource.attributes["host.id"]], "/")], ":")) where resource.attributes["cloud.resource_id"] == nil and resource.attributes["host.id"] != nil and resource.attributes["cloud.platform"] == "aws_ec2"
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], resource.attributes["ec2.tag.aws:autoscaling:groupName"]], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["ec2.tag.aws:autoscaling:groupName"] != nil
                 - set(resource.attributes["deployment.environment.name"], Concat([resource.attributes["cloud.platform"], "default"], ":")) where resource.attributes["deployment.environment.name"] == nil and resource.attributes["cloud.platform"] != nil
     transform/logs_cleanup: