Assume installed agent on upgrade

changelog/fragments/1730284359-assume-installed-agent-on-upgrade.yaml

@@ -0,0 +1,32 @@
+# Kind can be one of:
+# - breaking-change: a change to previously-documented behavior
+# - deprecation: functionality that is being removed in a later release
+# - bug-fix: fixes a problem in a previous version
+# - enhancement: extends functionality but does not break or fix existing behavior
+# - feature: new functionality
+# - known-issue: problems that we are aware of in a given version
+# - security: impacts on the security of a product or a user’s deployment.
+# - upgrade: important information for someone upgrading from a prior version
+# - other: does not fit into any of the other categories
+kind: bug-fix
+
+# Change summary; a 80ish characters long description of the change.
+summary: This PR assumes installed agent on upgrade in order to initialize control path properly.
+
+# Long description; in case the summary is not enough to describe the change
+# this field accommodate a description without length limits.
+# NOTE: This field will be rendered only for breaking-change and known-issue kinds at the moment.
+#description:
+
+# Affected component; usually one of "elastic-agent", "fleet-server", "filebeat", "metricbeat", "auditbeat", "all", etc.
+component: elastic-agent
+
+# PR URL; optional; the PR number that added the changeset.
+# If not present is automatically filled by the tooling finding the PR where this changelog fragment has been added.
+# NOTE: the tooling supports backports, so it's able to fill the original PR number instead of the backport PR number.
+# Please provide it if you are adding a fragment for a different PR.
+pr: https://github.com/owner/repo/5879
+
+# Issue URL; optional; the GitHub issue related to this changeset (either closes or is part of).
+# If not present is automatically filled by the tooling with the issue linked to the PR number.
+issue: https://github.com/owner/repo/5872

internal/pkg/agent/application/paths/paths_test.go

@@ -5,10 +5,12 @@
 package paths
 
 import (
+	"fmt"
 	"runtime"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )
 
 func TestEqual(t *testing.T) {
@@ -92,3 +94,123 @@ func TestHasPrefixWindows(t *testing.T) {
 		})
 	}
 }
+
+func TestResolveControlSocketWithInstalledState(t *testing.T) {
+	testCases := []struct {
+		os               string
+		controlSocket    string
+		topPath          string
+		runningInstalled bool
+		expectedSocket   string
+	}{
+		{
+			"darwin",
+			ControlSocketFromPath("darwin", "/top"),
+			"/top",
+			true,
+			ControlSocketFromPath("darwin", "/top"),
+		},
+		{
+			"darwin",
+			ControlSocketFromPath("darwin", "/top"),
+			"/top",
+			false,
+			ControlSocketFromPath("darwin", "/top"),
+		},
+		{
+			"darwin",
+			"/control/socket",
+			"/top",
+			true,
+			"/control/socket",
+		},
+		{
+			"darwin",
+			"/control/socket",
+			"/top",
+			false,
+			"/control/socket",
+		},
+
+		{
+			"linux",
+			ControlSocketFromPath("linux", "/top"),
+			"/top",
+			true,
+			ControlSocketFromPath("linux", "/top"),
+		},
+		{
+			"linux",
+			ControlSocketFromPath("linux", "/top"),
+			"/top",
+			false,
+			ControlSocketFromPath("linux", "/top"),
+		},
+		{
+			"linux",
+			"/control/socket",
+			"/top",
+			true,
+			"/control/socket",
+		},
+		{
+			"linux",
+			"/control/socket",
+			"/top",
+			false,
+			"/control/socket",
+		},
+
+		{
+			"windows",
+			ControlSocketFromPath("windows", "/top"),
+			"/top",
+			true,
+			WindowsControlSocketInstalledPath,
+		},
+		{
+			"windows",
+			ControlSocketFromPath("windows", "/top"),
+			"/top",
+			false,
+			ControlSocketFromPath("windows", "/top"),
+		},
+
+		{
+			"windows",
+			"/control/socket",
+			"/top",
+			true,
+			"/control/socket",
+		},
+		{
+			"windows",
+			"/control/socket",
+			"/top",
+			false,
+			"/control/socket",
+		},
+	}
+
+	for i, tc := range testCases {
+		if runtime.GOOS != tc.os {
+			continue
+		}
+		t.Run(fmt.Sprintf("test case %d", i), func(t *testing.T) {
+			prevControlSocketPath := controlSocketPath
+			prevTopPath := topPath
+			defer func() {
+				// just to be sure
+				controlSocketPath = prevControlSocketPath
+				topPath = prevTopPath
+			}()
+			controlSocketPath = tc.controlSocket
+			topPath = tc.topPath
+
+			ResolveControlSocketWithInstalledState(tc.runningInstalled)
+
+			require.Equal(t, tc.expectedSocket, controlSocketPath)
+		})
+
+	}
+}

internal/pkg/agent/application/paths/paths_unix.go

@@ -35,6 +35,9 @@ func initialControlSocketPath(topPath string) string {
 // ResolveControlSocket does nothing on non-Windows hosts.
 func ResolveControlSocket() {}
 
+// ResolveControlSocketWithInstalledState updates the control socket path.
+func ResolveControlSocketWithInstalledState(_ bool) {}
+
 // HasPrefix tests if the path starts with the prefix.
 func HasPrefix(path string, prefix string) bool {
 	if path == "" || prefix == "" {

internal/pkg/agent/application/paths/paths_windows.go

@@ -65,8 +65,13 @@ func initialControlSocketPath(topPath string) string {
 // RunningInstalled will always be false, even when it is an installed version. Once
 // that is fixed from the upgrade process the control socket path needs to be updated.
 func ResolveControlSocket() {
+	ResolveControlSocketWithInstalledState(RunningInstalled())
+}
+
+// ResolveControlSocketWithInstalledState updates the control socket path.
+func ResolveControlSocketWithInstalledState(runningInstalled bool) {
 	currentPath := ControlSocket()
-	if currentPath == ControlSocketFromPath(runtime.GOOS, topPath) && RunningInstalled() {
+	if currentPath == ControlSocketFromPath(runtime.GOOS, topPath) && runningInstalled {
 		// path is not correct being that it's installed
 		// reset the control socket path to be the installed path
 		SetControlSocket(WindowsControlSocketInstalledPath)

internal/pkg/agent/application/upgrade/watcher.go

@@ -12,6 +12,7 @@ import (
 
 	"google.golang.org/grpc"
 
+	"github.com/elastic/elastic-agent/internal/pkg/agent/application/paths"
 	"github.com/elastic/elastic-agent/pkg/control/v2/client"
 	"github.com/elastic/elastic-agent/pkg/core/logger"
 )
@@ -50,6 +51,11 @@ type AgentWatcher struct {
 
 // NewAgentWatcher creates a new agent watcher.
 func NewAgentWatcher(ch chan error, log *logger.Logger, checkInterval time.Duration) *AgentWatcher {
+	// when starting watcher from pre 8.8 version of agent control socket is evaluated incorrectly and upgrade fails.
+	// resolving control socket updates it to a proper value before client is initiated
+	// upgrade is only available for installed agent so we can assume
+	paths.ResolveControlSocketWithInstalledState(true)
+
 	c := client.New()
 	ec := &AgentWatcher{
 		notifyChan:    ch,