Make captured env vars configurable

cli_flags.go

@@ -26,6 +26,7 @@ const (
 	defaultProbabilisticInterval  = 1 * time.Minute
 	defaultArgSendErrorFrames     = false
 	defaultOffCPUThreshold        = 0
+	defaultEnvVarsValue           = ""
 
 	// This is the X in 2^(n + x) where n is the default hardcoded map size value
 	defaultArgMapScaleFactor = 0
@@ -67,6 +68,7 @@ var (
 		"Valid values are in the range [1..%d], and 0 to disable off-cpu profiling."+
 		"Default is %d.",
 		support.OffCPUThresholdMax, defaultOffCPUThreshold)
+	envVarsHelp = "Defines a comma seperated list of env vars that should be reported with the captured profiling samples"
 )
 
 // Package-scope variable, so that conditionally compiled other components can refer
@@ -123,6 +125,8 @@ func parseArgs() (*controller.Config, error) {
 	fs.UintVar(&args.OffCPUThreshold, "off-cpu-threshold",
 		defaultOffCPUThreshold, offCPUThresholdHelp)
 
+	fs.StringVar(&args.IncludeEnvVars, "env-vars", defaultEnvVarsValue, envVarsHelp)
+
 	fs.Usage = func() {
 		fs.PrintDefaults()
 	}

internal/controller/config.go

@@ -41,6 +41,8 @@ type Config struct {
 	Reporter reporter.Reporter
 
 	Fs *flag.FlagSet
+
+	IncludeEnvVars string
 }
 
 const (

internal/controller/controller.go

@@ -3,6 +3,7 @@ package controller // import "go.opentelemetry.io/ebpf-profiler/internal/control
 import (
 	"context"
 	"fmt"
+	"strings"
 	"time"
 
 	log "github.com/sirupsen/logrus"
@@ -71,6 +72,14 @@ func (c *Controller) Start(ctx context.Context) error {
 		return fmt.Errorf("failed to start reporter: %w", err)
 	}
 
+	var envVars map[string]bool
+	splittedEnvVars := strings.Split(c.config.IncludeEnvVars, ",")
+	for _, envVar := range splittedEnvVars {
+		if envVar != "" {
+			envVars[envVar] = true
+		}
+	}
+
 	metrics.SetReporter(c.reporter)
 
 	// Load the eBPF code and map definitions
@@ -87,6 +96,7 @@ func (c *Controller) Start(ctx context.Context) error {
 		ProbabilisticInterval:  c.config.ProbabilisticInterval,
 		ProbabilisticThreshold: c.config.ProbabilisticThreshold,
 		OffCPUThreshold:        uint32(c.config.OffCPUThreshold),
+		IncludeEnvVars:         envVars,
 	})
 	if err != nil {
 		return fmt.Errorf("failed to load eBPF tracer: %w", err)

processmanager/manager.go

@@ -66,7 +66,7 @@ var (
 // implementation.
 func New(ctx context.Context, includeTracers types.IncludedTracers, monitorInterval time.Duration,
 	ebpf pmebpf.EbpfHandler, fileIDMapper FileIDMapper, symbolReporter reporter.SymbolReporter,
-	sdp nativeunwind.StackDeltaProvider, filterErrorFrames bool) (*ProcessManager, error) {
+	sdp nativeunwind.StackDeltaProvider, filterErrorFrames bool, includeEnvVars map[string]bool) (*ProcessManager, error) {
 	if fileIDMapper == nil {
 		var err error
 		fileIDMapper, err = newFileIDMapper(lruFileIDCacheSize)
@@ -101,6 +101,7 @@ func New(ctx context.Context, includeTracers types.IncludedTracers, monitorInter
 		reporter:                 symbolReporter,
 		metricsAddSlice:          metrics.AddSlice,
 		filterErrorFrames:        filterErrorFrames,
+		includeEnvVars:           includeEnvVars,
 	}
 
 	collectInterpreterMetrics(ctx, pm, monitorInterval)

processmanager/processinfo.go

@@ -16,7 +16,6 @@ import (
 	"fmt"
 	"os"
 	"path"
-	"slices"
 	"strings"
 	"syscall"
 	"time"
@@ -38,15 +37,6 @@ import (
 	"go.opentelemetry.io/ebpf-profiler/util"
 )
 
-type ProcessManagerConfig struct {
-	extractEnvVars []string
-}
-
-var pm_cfg = ProcessManagerConfig{extractEnvVars: []string{
-	"PIPELINE_PPOID",
-	"PIPELINE_SOFTWARENAME",
-	"PIPELINE_JOBID"}}
-
 // assignTSDInfo updates the TSDInfo for the Interpreters on given PID.
 // Caller must hold pm.mu write lock.
 func (pm *ProcessManager) assignTSDInfo(pid libpf.PID, tsdInfo *tpbase.TSDInfo) {
@@ -105,7 +95,9 @@ func (pm *ProcessManager) updatePidInformation(pid libpf.PID, m *Mapping) (bool,
 			splittedVars := strings.Split(string(envVars), "\000")
 			for _, envVar := range splittedVars {
 				keyValuePair := strings.SplitN(envVar, "=", 2)
-				if slices.Contains(pm_cfg.extractEnvVars, keyValuePair[0]) {
+				_, envVarShouldBeCaptured := pm.includeEnvVars[keyValuePair[0]]
+
+				if envVarShouldBeCaptured {
 					envVarMap[keyValuePair[0]] = keyValuePair[1]
 				}
 			}

processmanager/types.go

@@ -96,6 +96,9 @@ type ProcessManager struct {
 
 	// filterErrorFrames determines whether error frames are dropped by `ConvertTrace`.
 	filterErrorFrames bool
+
+	// includeEnvVars holds a list of env vars that should be captured from processes
+	includeEnvVars map[string]bool
 }
 
 // Mapping represents an executable memory mapping of a process.

tracer/tracer.go

@@ -156,6 +156,8 @@ type Config struct {
 	ProbabilisticThreshold uint
 	// OffCPUThreshold is the user defined threshold for off-cpu profiling.
 	OffCPUThreshold uint32
+	// IncludeEnvVars holds a list of environment variables that should be captured and reported from processes
+	IncludeEnvVars map[string]bool
 }
 
 // hookPoint specifies the group and name of the hooked point in the kernel.
@@ -296,7 +298,7 @@ func NewTracer(ctx context.Context, cfg *Config) (*Tracer, error) {
 
 	processManager, err := pm.New(ctx, cfg.IncludeTracers, cfg.Intervals.MonitorInterval(),
 		ebpfHandler, nil, cfg.Reporter, elfunwindinfo.NewStackDeltaProvider(),
-		cfg.FilterErrorFrames)
+		cfg.FilterErrorFrames, cfg.IncludeEnvVars)
 	if err != nil {
 		return nil, fmt.Errorf("failed to create processManager: %v", err)
 	}