Files
vantage-agent/internal/sync/sync.go
T
mrhid6 b5b9775d2b
Agent Release / build (push) Successful in 3m40s
Agent Release / msi (push) Successful in 4m54s
fix(patch): gate phases on the window deadline, queue undelivered results, retry startup inventory
The window deadline no longer kills a running package manager: it only gates
the start of each phase, and a started upgrade runs under a 2 hour backstop
that sends SIGTERM on Linux. A PatchResult whose send fails is queued and
flushed on the next command stream, retaking the reboot decision. deb822
folded Suites continuation lines are filtered with the field. The startup
static inventory report is retried until it succeeds.
2026-09-15 13:43:03 +00:00

812 lines
24 KiB
Go

package agentsync
import (
"context"
"crypto/sha256"
"encoding/hex"
"fmt"
"io"
"log"
"net"
"net/http"
"os"
"os/exec"
"path/filepath"
"runtime"
"strings"
"sync"
"time"
"gitea.hostxtra.co.uk/vantage/vantage-agent/internal/config"
agentexec "gitea.hostxtra.co.uk/vantage/vantage-agent/internal/exec"
grpcclient "gitea.hostxtra.co.uk/vantage/vantage-agent/internal/grpc"
"gitea.hostxtra.co.uk/vantage/vantage-agent/internal/inventory"
"gitea.hostxtra.co.uk/vantage/vantage-agent/internal/keys"
"gitea.hostxtra.co.uk/vantage/vantage-agent/internal/monitors"
agentproxy "gitea.hostxtra.co.uk/vantage/vantage-agent/internal/proxy"
"gitea.hostxtra.co.uk/vantage/vantage-agent/internal/updates"
"gitea.hostxtra.co.uk/vantage/vantage-shared/grpc/pb"
)
func Run(ctx context.Context, cfg *config.Config, version string) error {
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
return fmt.Errorf("dial grpc: %w", err)
}
defer client.Close()
if cfg.PreRegToken != "" {
log.Println("registering with server...")
hostname, _ := os.Hostname()
ipAddress := localIP()
osInfo := fmt.Sprintf("%s %s", runtime.GOOS, runtime.GOARCH)
agentToken, err := client.Register(cfg.ServerID, cfg.PreRegToken, hostname, ipAddress, osInfo)
if err != nil {
return fmt.Errorf("registration failed: %w", err)
}
cfg.AgentToken = agentToken
cfg.PreRegToken = ""
if err := config.Save(cfg); err != nil {
return fmt.Errorf("save config: %w", err)
}
log.Println("registration successful")
client.Close()
client, err = grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
return fmt.Errorf("reconnect: %w", err)
}
}
if cfg.AgentToken == "" {
return fmt.Errorf("no agent token available registration required")
}
go runCommandStream(ctx, cfg)
go runUpdateCheck(ctx, cfg)
go runInventory(ctx, cfg)
go runWorkloads(ctx, cfg)
go monitors.Run(ctx, cfg)
ticker := time.NewTicker(cfg.PollInterval)
defer ticker.Stop()
if err := poll(client, cfg, version); err != nil {
log.Printf("poll error: %v", err)
}
for {
select {
case <-ctx.Done():
return nil
case <-ticker.C:
if err := poll(client, cfg, version); err != nil {
log.Printf("poll error: %v", err)
}
}
}
}
func poll(client *grpcclient.Client, cfg *config.Config, version string) error {
resp, err := client.SyncKeys(cfg.ServerID, cfg.AgentToken, version)
if err != nil {
return fmt.Errorf("SyncKeys: %w", err)
}
// Stored atomically: the hourly package loop reads this from another
// goroutine. Absent on the wire decodes as false, so an older server leaves
// collection off rather than on.
collectPackagesFlag.Store(resp.CollectPackages)
markFirstPoll()
desired := resp.PublicKeys
if runtime.GOOS != "linux" {
return nil
}
current, err := keys.ReadAuthorizedKeys()
if err != nil {
return fmt.Errorf("read authorized_keys: %w", err)
}
if !keys.StateChanged(current, desired) {
log.Println("authorized_keys unchanged, skipping write")
return nil
}
if err := keys.WriteAuthorizedKeys(desired); err != nil {
return fmt.Errorf("write authorized_keys: %w", err)
}
log.Printf("authorized_keys updated (%d keys)", len(desired))
return nil
}
// How long a command stream must survive before it counts as having worked.
// Past this, the next drop is treated as a fresh incident rather than as the
// continuation of a run of failures.
const streamHealthyAfter = time.Minute
// Stream staleness. The server beats every 20s, so 70s tolerates three missed
// beats before the stream is written off - high enough that a slow network or a
// briefly busy server does not cost a reconnect, low enough that an agent is
// not uncommandable for minutes after a control-plane restart.
const (
streamStaleAfter = 70 * time.Second
streamStaleCheck = 10 * time.Second
// How often a healthy stream reports itself. Also the interval at which an
// agent talking to a control plane too old to send heartbeats says so -
// that agent is running without a watchdog, and the journal should not be
// silent about it.
pingSummaryInterval = 5 * time.Minute
)
func runCommandStream(ctx context.Context, cfg *config.Config) {
backoff := time.Second
// Two minutes was the old ceiling, and it was reached far too easily. The
// command stream is what makes this agent controllable at all: while it is
// down, workflows and console sessions fail as "agent offline" even though
// SyncKeys keeps polling happily and the fleet list still shows the server
// active. A shorter ceiling costs a few reconnect attempts; the old one cost
// two minutes of an agent that looks fine and answers nothing.
const maxBackoff = 30 * time.Second
for {
select {
case <-ctx.Done():
return
default:
}
started := time.Now()
err := connectAndHandleStream(ctx, cfg)
if ctx.Err() != nil {
return
}
// A stream that stayed up is evidence the control plane is reachable,
// whatever ended it. Without this the backoff only ever climbed:
// connectAndHandleStream returns an error on *every* stream end,
// including a healthy one dropped by a routine deploy, so an agent
// pinned itself at the ceiling after a handful of ordinary restarts and
// stayed there for the rest of its life.
if time.Since(started) >= streamHealthyAfter {
backoff = time.Second
}
// The uptime is in the line because it is what distinguishes a stream
// that never worked from one that ran for hours and was dropped by a
// deploy - and it is the same measure that decides whether the backoff
// resets, so a reader can see why the delay is what it is.
up := time.Since(started).Truncate(time.Second)
if err != nil {
log.Printf("command stream error after %s: %v, reconnecting in %s", up, err, backoff)
} else {
log.Printf("command stream closed after %s, reconnecting in %s", up, backoff)
}
select {
case <-ctx.Done():
return
case <-time.After(backoff):
}
if backoff < maxBackoff {
backoff *= 2
if backoff > maxBackoff {
backoff = maxBackoff
}
}
}
}
func connectAndHandleStream(ctx context.Context, cfg *config.Config) error {
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
return fmt.Errorf("dial: %w", err)
}
defer client.Close()
// Cancelling this context is what unblocks Recv when the stream has gone
// quiet. Without it the watchdog below would have no way to interrupt a
// read that is never going to return.
streamCtx, abandon := context.WithCancel(ctx)
defer abandon()
stream, err := client.CommandStream(streamCtx)
if err != nil {
return fmt.Errorf("open stream: %w", err)
}
if err := stream.Send(&pb.AgentMessage{
ServerId: cfg.ServerID,
AgentToken: cfg.AgentToken,
Ready: &pb.AgentReady{},
}); err != nil {
return fmt.Errorf("send auth: %w", err)
}
log.Printf("command stream connected to %s", cfg.ServerURL)
var sendMu sync.Mutex
send := func(msg *pb.AgentMessage) error {
sendMu.Lock()
defer sendMu.Unlock()
return stream.Send(msg)
}
// Results that could not be sent on an earlier stream go out first. In
// its own goroutine: a queued reboot re-check can take a while on
// Windows, and the receive loop below must start promptly.
go flushPendingResults(send)
// Stream liveness, tracked here rather than left to gRPC keepalive.
//
// Keepalive operates on the transport, and behind an L7 proxy the transport
// ends at the proxy: it answers pings whether or not the server behind it
// is still running. A control-plane pod that dies therefore leaves this
// agent blocked in Recv on a stream that will never deliver another message
// and never error, while the control plane dispatches commands into it and
// the operator watches nothing happen.
//
// The watchdog only arms once a ping has actually been seen. A server too
// old to send them must not be treated as dead - that would put the agent
// in a reconnect loop against a control plane that is working perfectly.
var (
lastMu sync.Mutex
lastRecv = time.Now()
pinged bool
beats int
)
markRecv := func(isPing bool) {
lastMu.Lock()
lastRecv = time.Now()
if isPing {
beats++
// Logged once per stream, because it is the moment the agent starts
// holding the control plane to account: before this the watchdog is
// disarmed and a dead stream would go unnoticed indefinitely.
if !pinged {
pinged = true
log.Printf("command stream heartbeat detected, watchdog armed (%s threshold)", streamStaleAfter)
}
}
lastMu.Unlock()
}
go func() {
t := time.NewTicker(streamStaleCheck)
defer t.Stop()
// Reported periodically rather than per beat: at one every 20s the
// journal would be nothing else. The count is what makes a partial
// failure visible - beats arriving but fewer than expected is a
// different problem from beats stopping altogether.
summary := time.NewTicker(pingSummaryInterval)
defer summary.Stop()
for {
select {
case <-streamCtx.Done():
return
case <-summary.C:
lastMu.Lock()
n, armed := beats, pinged
beats = 0
lastMu.Unlock()
if armed {
log.Printf("command stream healthy, %d heartbeats in the last %s", n, pingSummaryInterval)
} else {
log.Printf("command stream up but sending no heartbeats; "+
"control plane predates them, watchdog stays disarmed (last message %s ago)",
time.Since(lastRecv).Truncate(time.Second))
}
case <-t.C:
lastMu.Lock()
idle, armed := time.Since(lastRecv), pinged
lastMu.Unlock()
if armed && idle > streamStaleAfter {
log.Printf("command stream silent for %s (threshold %s), assuming it is dead and reconnecting",
idle.Truncate(time.Second), streamStaleAfter)
abandon()
return
}
}
}
}()
for {
cmd, err := stream.Recv()
if err != nil {
return fmt.Errorf("recv: %w", err)
}
markRecv(cmd.Ping != nil)
// Pings carry nothing and are not acknowledged; being received is their
// whole purpose.
if cmd.Ping != nil {
continue
}
if cmd.GenerateKey != nil {
go handleGenerateKey(cfg, cmd)
}
if cmd.DeleteKey != nil {
go handleDeleteKey(cmd)
}
if cmd.UpdateAgent != nil {
go handleUpdateAgent(cmd)
}
if cmd.ApplyUpdates != nil {
go handleApplyUpdates(send, cfg, cmd)
}
if cmd.CleanupWorkspace != nil {
go handleCleanupWorkspace(cmd)
}
if cmd.OpenProxy != nil {
go handleOpenProxy(ctx, cfg, cmd.OpenProxy)
}
if cmd.RefreshWorkloads != nil {
go handleRefreshWorkloads(cfg)
}
if cmd.ControlWorkload != nil {
go handleControlWorkload(send, cfg, cmd.CommandId, cmd.ControlWorkload)
}
if cmd.WorkloadLogs != nil {
go handleWorkloadLogs(send, cfg, cmd.CommandId, cmd.WorkloadLogs)
}
if cmd.RunStep != nil {
go func(rc *pb.RunStepCmd, cid string) {
emit := func(seq uint64, data []byte) {
_ = send(&pb.AgentMessage{
ServerId: cfg.ServerID,
AgentToken: cfg.AgentToken,
StepOutput: &pb.StepOutputChunk{CommandId: cid, Seq: seq, Data: data},
})
}
res := agentexec.RunStep(rc, emit)
res.CommandId = cid
_ = send(&pb.AgentMessage{
ServerId: cfg.ServerID,
AgentToken: cfg.AgentToken,
StepOutput: &pb.StepOutputChunk{CommandId: cid, Eof: true},
})
_ = send(&pb.AgentMessage{
ServerId: cfg.ServerID,
AgentToken: cfg.AgentToken,
StepResult: res,
})
}(cmd.RunStep, cmd.CommandId)
continue
}
}
}
func runUpdateCheck(ctx context.Context, cfg *config.Config) {
const interval = time.Hour
doCheck := func() {
pkgs, err := updates.CheckAvailable()
if err != nil {
log.Printf("update check error: %v", err)
return
}
pbUpdates := make([]pb.PackageUpdate, len(pkgs))
for i, p := range pkgs {
pbUpdates[i] = pb.PackageUpdate{
Name: p.Name,
CurrentVersion: p.CurrentVersion,
NewVersion: p.NewVersion,
}
}
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
log.Printf("update report dial error: %v", err)
return
}
defer client.Close()
if err := client.ReportUpdates(cfg.ServerID, cfg.AgentToken, pbUpdates); err != nil {
log.Printf("ReportUpdates error: %v", err)
return
}
log.Printf("reported %d available OS updates", len(pkgs))
// Same hourly cadence, same connection. A package set changes on
// roughly the schedule available updates do, so this needs no timer of
// its own.
reportPackages(client, cfg)
}
// The boot round only: after this the flag has long been set, and every
// later tick is an hour past a poll that runs every 30s.
waitFirstPoll(ctx, firstPollWait)
doCheck()
ticker := time.NewTicker(interval)
defer ticker.Stop()
for {
select {
case <-ctx.Done():
return
case <-ticker.C:
doCheck()
}
}
}
func runInventory(ctx context.Context, cfg *config.Config) {
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
log.Printf("inventory dial error: %v", err)
return
}
defer client.Close()
report := func(static bool) error {
r := inventory.Collect(static)
r.ServerId = cfg.ServerID
r.AgentToken = cfg.AgentToken
// Static snapshots only - every 15 minutes, not every 30 seconds. On
// Windows this spawns a PowerShell process, which is not something to
// do twice a minute forever, and a host rebooted by hand clearing the
// flag within a quarter of an hour is soon enough.
//
// Computed here rather than inside inventory.Collect so the inventory
// package gains no dependency on updates.
if static {
r.RebootRequired = updates.RebootRequired()
}
if err := client.ReportInventory(r); err != nil {
log.Printf("report inventory: %v", err)
return err
}
return nil
}
// The startup static report carries the boot time the server uses to
// prove a patch reboot happened, so a failure is retried on the next
// ticks (every 30 seconds, up to startupStaticAttempts in total) instead
// of waiting a quarter of an hour for the next static snapshot.
const startupStaticAttempts = 10
attempts := 1
startupPending := report(true) != nil
ticker := time.NewTicker(30 * time.Second)
defer ticker.Stop()
tick := 0
for {
select {
case <-ctx.Done():
return
case <-ticker.C:
tick++
if startupPending && attempts < startupStaticAttempts {
attempts++
startupPending = report(true) != nil
continue
}
startupPending = false
report(tick%30 == 0)
}
}
}
func handleCleanupWorkspace(cmd *pb.ServerCommand) {
id := cmd.CleanupWorkspace.WorkspaceId
dir := agentexec.WorkspacePath(id)
if err := os.RemoveAll(dir); err != nil {
log.Printf("cleanup workspace %s failed (cmd=%s): %v", dir, cmd.CommandId, err)
return
}
log.Printf("removed run workspace %s (cmd=%s)", dir, cmd.CommandId)
}
// handleOpenProxy relays one console connection. It uses its own gRPC
// connection so console traffic never shares a stream with commands, key sync
// or workflow output.
func handleOpenProxy(ctx context.Context, cfg *config.Config, cmd *pb.OpenProxyCmd) {
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
log.Printf("proxy %s: dial control plane: %v", cmd.ProxyId, err)
return
}
defer client.Close()
stream, err := client.ProxyStream(ctx)
if err != nil {
log.Printf("proxy %s: open stream: %v", cmd.ProxyId, err)
return
}
log.Printf("proxy %s: relaying 127.0.0.1:%d", cmd.ProxyId, cmd.Port)
if err := agentproxy.Open(stream, cfg.ServerID, cfg.AgentToken, cmd.ProxyId, cmd.Port); err != nil {
log.Printf("proxy %s: %v", cmd.ProxyId, err)
}
}
func handleDeleteKey(cmd *pb.ServerCommand) {
label := cmd.DeleteKey.Label
keyPath := fmt.Sprintf("/root/.ssh/vantage_%s", strings.ReplaceAll(label, " ", "_"))
if err := keys.RemoveSSHIdentity(keyPath); err != nil {
log.Printf("remove ssh identity failed (cmd=%s): %v", cmd.CommandId, err)
}
for _, path := range []string{keyPath, keyPath + ".pub"} {
if err := os.Remove(path); err != nil && !os.IsNotExist(err) {
log.Printf("delete key file %s (cmd=%s): %v", path, cmd.CommandId, err)
}
}
log.Printf("deleted local key files for %q (cmd=%s)", label, cmd.CommandId)
}
func handleUpdateAgent(cmd *pb.ServerCommand) {
if runtime.GOOS == "windows" {
handleUpdateAgentWindows(cmd)
return
}
u := cmd.UpdateAgent
arch := runtime.GOARCH
tag := "agent%2Fv" + u.Version
binaryURL := fmt.Sprintf("%s/vantage/vantage-agent/releases/download/%s/vantage-agent-linux-%s", u.GiteaBaseURL, tag, arch)
checksumURL := fmt.Sprintf("%s/vantage/vantage-agent/releases/download/%s/checksums.txt", u.GiteaBaseURL, tag)
log.Printf("updating agent to v%s from %s (cmd=%s)", u.Version, u.GiteaBaseURL, cmd.CommandId)
tmpBin := "/tmp/vantage-agent-update"
if err := downloadFile(binaryURL, tmpBin); err != nil {
log.Printf("update download failed (cmd=%s): %v", cmd.CommandId, err)
return
}
checksumData, err := httpGetBytes(checksumURL)
if err != nil {
log.Printf("update checksum fetch failed (cmd=%s): %v", cmd.CommandId, err)
return
}
if err := verifyChecksum(tmpBin, fmt.Sprintf("vantage-agent-linux-%s", arch), checksumData); err != nil {
log.Printf("update checksum mismatch (cmd=%s): %v", cmd.CommandId, err)
os.Remove(tmpBin)
return
}
if err := os.Chmod(tmpBin, 0755); err != nil {
log.Printf("update chmod failed (cmd=%s): %v", cmd.CommandId, err)
return
}
if err := os.Rename(tmpBin, "/usr/local/bin/vantage-agent"); err != nil {
log.Printf("update replace binary failed (cmd=%s): %v", cmd.CommandId, err)
return
}
log.Printf("agent binary replaced, restarting service (cmd=%s)", cmd.CommandId)
exec.Command("systemctl", "restart", "vantage-agent").Run()
}
func handleUpdateAgentWindows(cmd *pb.ServerCommand) {
u := cmd.UpdateAgent
tag := "agent%2Fv" + u.Version
msiURL := fmt.Sprintf("%s/vantage/vantage-agent/releases/download/%s/vantage-agent.msi", u.GiteaBaseURL, tag)
checksumURL := fmt.Sprintf("%s/vantage/vantage-agent/releases/download/%s/checksums-msi.txt", u.GiteaBaseURL, tag)
log.Printf("updating agent to v%s from %s (cmd=%s)", u.Version, u.GiteaBaseURL, cmd.CommandId)
msiPath := filepath.Join(os.TempDir(), "vantage-agent-update.msi")
if err := downloadFile(msiURL, msiPath); err != nil {
log.Printf("update download failed (cmd=%s): %v", cmd.CommandId, err)
return
}
checksumData, err := httpGetBytes(checksumURL)
if err != nil {
log.Printf("update checksum fetch failed (cmd=%s): %v", cmd.CommandId, err)
return
}
if err := verifyChecksum(msiPath, "vantage-agent.msi", checksumData); err != nil {
log.Printf("update checksum mismatch (cmd=%s): %v", cmd.CommandId, err)
os.Remove(msiPath)
return
}
logPath := filepath.Join(os.TempDir(), "vantage-agent-msi.log")
// The MSI stops the vantage-agent service as part of the upgrade. Anything
// descended from this process is killed with it, so msiexec must not be a
// child: run it from a scheduled task, which is parented to the Task
// Scheduler service instead.
if err := launchDetachedUpdate(msiPath, logPath, cmd.CommandId); err != nil {
log.Printf("failed to launch msiexec (cmd=%s): %v", cmd.CommandId, err)
return
}
log.Printf("scheduled msiexec for upgrade to v%s (cmd=%s)", u.Version, cmd.CommandId)
}
const updateTaskName = "VantageAgentUpdate"
func launchDetachedUpdate(msiPath, logPath, commandID string) error {
scriptPath := filepath.Join(os.TempDir(), "vantage-agent-update.cmd")
script := fmt.Sprintf("@echo off\r\n"+
"timeout /t 5 /nobreak >nul\r\n"+
"msiexec /i \"%s\" /qn /norestart /l*v \"%s\"\r\n"+
"schtasks /delete /tn %s /f >nul 2>&1\r\n"+
"del /f /q \"%s\" >nul 2>&1\r\n"+
"(goto) 2>nul & del /f /q \"%%~f0\"\r\n",
msiPath, logPath, updateTaskName, msiPath)
if err := os.WriteFile(scriptPath, []byte(script), 0o600); err != nil {
return fmt.Errorf("write update script: %w", err)
}
// Stale task from a previous attempt would make /create fail even with /f
// if it is still running, so tear it down first and ignore the result.
exec.Command("schtasks", "/end", "/tn", updateTaskName).Run()
exec.Command("schtasks", "/delete", "/tn", updateTaskName, "/f").Run()
create := exec.Command("schtasks", "/create",
"/tn", updateTaskName,
"/tr", `"`+scriptPath+`"`,
"/sc", "once",
// Already in the past: the task never fires on its own, only via /run.
"/st", "00:00",
"/ru", "SYSTEM",
"/rl", "HIGHEST",
"/f")
if out, err := create.CombinedOutput(); err != nil {
return fmt.Errorf("schtasks create: %v: %s", err, strings.TrimSpace(string(out)))
}
if out, err := exec.Command("schtasks", "/run", "/tn", updateTaskName).CombinedOutput(); err != nil {
return fmt.Errorf("schtasks run: %v: %s", err, strings.TrimSpace(string(out)))
}
return nil
}
func downloadFile(url, dest string) error {
resp, err := http.Get(url)
if err != nil {
return err
}
defer resp.Body.Close()
if resp.StatusCode != http.StatusOK {
return fmt.Errorf("HTTP %d from %s", resp.StatusCode, url)
}
f, err := os.Create(dest)
if err != nil {
return err
}
defer f.Close()
_, err = io.Copy(f, resp.Body)
return err
}
func httpGetBytes(url string) ([]byte, error) {
resp, err := http.Get(url)
if err != nil {
return nil, err
}
defer resp.Body.Close()
if resp.StatusCode != http.StatusOK {
return nil, fmt.Errorf("HTTP %d from %s", resp.StatusCode, url)
}
return io.ReadAll(resp.Body)
}
func verifyChecksum(filePath, filename string, checksumData []byte) error {
f, err := os.Open(filePath)
if err != nil {
return err
}
defer f.Close()
h := sha256.New()
if _, err := io.Copy(h, f); err != nil {
return err
}
actual := hex.EncodeToString(h.Sum(nil))
for _, line := range strings.Split(string(checksumData), "\n") {
fields := strings.Fields(line)
if len(fields) == 2 && fields[1] == filename {
if fields[0] != actual {
return fmt.Errorf("expected %s got %s", fields[0], actual)
}
return nil
}
}
return fmt.Errorf("no checksum entry found for %s", filename)
}
func handleGenerateKey(cfg *config.Config, cmd *pb.ServerCommand) {
g := cmd.GenerateKey
label := g.Label
keyPath := fmt.Sprintf("/root/.ssh/vantage_%s", strings.ReplaceAll(label, " ", "_"))
opts := keys.KeyGenOptions{
KeyType: g.KeyType,
KeySize: g.KeySize,
Passphrase: g.Passphrase,
Comment: g.Comment,
}
pubKey, err := keys.GenerateKeyPair(keyPath, opts)
if err != nil {
log.Printf("key generation failed (cmd=%s): %v", cmd.CommandId, err)
return
}
privKeyData, err := os.ReadFile(keyPath)
if err != nil {
log.Printf("read private key failed (cmd=%s): %v", cmd.CommandId, err)
return
}
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
log.Printf("dial for key upload failed (cmd=%s): %v", cmd.CommandId, err)
return
}
defer client.Close()
keyID, err := client.UploadGeneratedKey(cfg.ServerID, cfg.AgentToken, pubKey, string(privKeyData), label)
if err != nil {
log.Printf("key upload failed (cmd=%s): %v", cmd.CommandId, err)
return
}
if err := keys.AddSSHIdentity(keyPath); err != nil {
log.Printf("add ssh identity failed (cmd=%s): %v", cmd.CommandId, err)
}
log.Printf("generated and uploaded key %q (key_id=%s, cmd=%s)", label, keyID, cmd.CommandId)
}
func localIP() string {
addrs, err := net.InterfaceAddrs()
if err != nil {
return ""
}
for _, addr := range addrs {
if ipNet, ok := addr.(*net.IPNet); ok && !ipNet.IP.IsLoopback() {
if ipNet.IP.To4() != nil {
return ipNet.IP.String()
}
}
}
return ""
}
func GenerateAndUpload(cfg *config.Config, label string) error {
client, err := grpcclient.New(cfg.ServerURL, cfg.TLS)
if err != nil {
return err
}
defer client.Close()
keyPath := fmt.Sprintf("/root/.ssh/vantage_%s", strings.ReplaceAll(label, " ", "_"))
pubKey, err := keys.GenerateKeyPair(keyPath, keys.KeyGenOptions{Comment: label})
if err != nil {
return err
}
privKeyData, err := os.ReadFile(keyPath)
if err != nil {
return fmt.Errorf("read private key: %w", err)
}
keyID, err := client.UploadGeneratedKey(cfg.ServerID, cfg.AgentToken, pubKey, string(privKeyData), label)
if err != nil {
return err
}
if err := keys.AddSSHIdentity(keyPath); err != nil {
log.Printf("add ssh identity: %v", err)
}
log.Printf("uploaded generated key %s (key_id=%s)", label, keyID)
return nil
}