sendnrw/netbird
2
0
Fork 0
mirror of https://github.com/netbirdio/netbird.git synced 2026-10-01 11:09:15 +02:00
Code Issues Packages Projects Releases Wiki Activity

[client] Add debug cpu start and stop commands (#7749)

Browse Source 
* Add debug cpu start and stop commands to profile the daemon without a restart

* Restore test globals on every exit and stop the daemon in the cpu profile test

* Add a no-updown flag to debug for

* Enable sync response persistence with --no-updown and reset flags between debug test runs

* Reset flags of every command between debug test runs

* Reset slice flags with Replace in the debug test helper

* Explain a running CPU profile in debug for and document cpu start and no-updown limits
This commit is contained in:
Viktor Liu
2026-09-30 17:37:49 +02:00
committed by GitHub
parent 0dc729c4ea
commit 6c453a0f97
3 changed files with 305 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files
client/cmd/debug.go
+58 -26
View File
@@ -23,7 +23,10 @@ import (
"github.com/netbirdio/netbird/version"
)
const errCloseConnection = "Failed to close connection: %v"
const (
errCloseConnection = "Failed to close connection: %v"
noUpDownFlag = "no-updown"
)
var (
logFileCount uint32
@@ -257,13 +260,14 @@ func runForDuration(cmd *cobra.Command, args []string) error {
}
stateWasDown := stat.Status != string(internal.StatusConnected) && stat.Status != string(internal.StatusConnecting)
noUpDown, _ := cmd.Flags().GetBool(noUpDownFlag)
initialLogLevel, err := client.GetLogLevel(cmd.Context(), &proto.GetLogLevelRequest{})
if err != nil {
return fmt.Errorf("failed to get log level: %v", status.Convert(err).Message())
}
if stateWasDown {
if stateWasDown && !noUpDown {
if _, err := client.Up(cmd.Context(), &proto.UpRequest{}); err != nil {
cmd.PrintErrf("Failed to bring service up: %v\n", status.Convert(err).Message())
} else {
@@ -284,34 +288,20 @@ func runForDuration(cmd *cobra.Command, args []string) error {
}
needsRestoreUp := false
if _, err := client.Down(cmd.Context(), &proto.DownRequest{}); err != nil {
cmd.PrintErrf("Failed to bring service down: %v\n", status.Convert(err).Message())
if noUpDown {
enableSyncResponsePersistence(cmd, client)
} else {
needsRestoreUp = !stateWasDown
cmd.Println("netbird down")
needsRestoreUp = restartDaemon(cmd, client, stateWasDown)
}
time.Sleep(1 * time.Second)
// Enable sync response persistence before bringing the service up
if _, err := client.SetSyncResponsePersistence(cmd.Context(), &proto.SetSyncResponsePersistenceRequest{
Enabled: true,
}); err != nil {
cmd.PrintErrf("Failed to enable sync response persistence: %v\n", status.Convert(err).Message())
}
if _, err := client.Up(cmd.Context(), &proto.UpRequest{}); err != nil {
cmd.PrintErrf("Failed to bring service up: %v\n", status.Convert(err).Message())
} else {
needsRestoreUp = false
cmd.Println("netbird up")
}
time.Sleep(3 * time.Second)
cpuProfilingStarted := false
if _, err := client.StartCPUProfile(cmd.Context(), &proto.StartCPUProfileRequest{}); err != nil {
cmd.PrintErrf("Failed to start CPU profiling: %v\n", err)
if msg := status.Convert(err).Message(); strings.Contains(msg, "already in progress") {
cmd.PrintErrln("CPU profiling is already running (started with `netbird debug cpu start`). " +
"It is left running and is included in a bundle created after `netbird debug cpu stop`.")
} else {
cmd.PrintErrf("Failed to start CPU profiling: %v\n", msg)
}
} else {
cpuProfilingStarted = true
defer func() {
@@ -401,7 +391,7 @@ func runForDuration(cmd *cobra.Command, args []string) error {
}
}
if stateWasDown {
if stateWasDown && !noUpDown {
if _, err := client.Down(cmd.Context(), &proto.DownRequest{}); err != nil {
cmd.PrintErrf("Failed to restore service down state: %v\n", status.Convert(err).Message())
} else {
@@ -458,6 +448,47 @@ func setSyncResponsePersistence(cmd *cobra.Command, args []string) error {
return nil
}
// enableSyncResponsePersistence asks the daemon to keep the latest sync
// response so the bundle carries the network map. With a running daemon only
// syncs received after the call are kept.
func enableSyncResponsePersistence(cmd *cobra.Command, client proto.DaemonServiceClient) {
if _, err := client.SetSyncResponsePersistence(cmd.Context(), &proto.SetSyncResponsePersistenceRequest{
Enabled: true,
}); err != nil {
cmd.PrintErrf("Failed to enable sync response persistence: %v\n", status.Convert(err).Message())
}
}
// restartDaemon cycles the daemon down and up with sync response persistence
// enabled so the bundle carries the network map. It reports whether the
// daemon was left down although it was running before, so the caller can
// bring it back up.
func restartDaemon(cmd *cobra.Command, client proto.DaemonServiceClient, stateWasDown bool) bool {
needsRestoreUp := false
if _, err := client.Down(cmd.Context(), &proto.DownRequest{}); err != nil {
cmd.PrintErrf("Failed to bring service down: %v\n", status.Convert(err).Message())
} else {
needsRestoreUp = !stateWasDown
cmd.Println("netbird down")
}
time.Sleep(1 * time.Second)
// Enable sync response persistence before bringing the service up
enableSyncResponsePersistence(cmd, client)
if _, err := client.Up(cmd.Context(), &proto.UpRequest{}); err != nil {
cmd.PrintErrf("Failed to bring service up: %v\n", status.Convert(err).Message())
} else {
needsRestoreUp = false
cmd.Println("netbird up")
}
time.Sleep(3 * time.Second)
return needsRestoreUp
}
func waitForDurationOrCancel(ctx context.Context, duration time.Duration, cmd *cobra.Command) error {
ticker := time.NewTicker(1 * time.Second)
defer ticker.Stop()
@@ -546,4 +577,5 @@ func init() {
forCmd.Flags().StringVar(&uploadBundleURLFlag, "upload-bundle-url", types.DefaultBundleURL, "Service URL to get an URL to upload the debug bundle")
forCmd.Flags().BoolVar(&uploadBundleInsecureFlag, "upload-bundle-insecure", false, "Allow uploading to an http or untrusted-TLS upload server (self-hosted); requires root")
forCmd.Flags().Bool("capture", false, "Capture packets during the debug duration and include in bundle")
forCmd.Flags().Bool(noUpDownFlag, false, "Keep the daemon running instead of bringing it down and up before collecting. The bundle only includes the network map if a sync arrives during the run")
}