Skip to content

[release-4.22] OCPBUGS-109638: Fix clock_state metric: HOLDOVER mapped to 0, stale when CEP enabled #636

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jzding wants to merge 1 commit into
base: release-4.22
Choose a base branch
from
+63 −22
Open

[release-4.22] OCPBUGS-109638: Fix clock_state metric: HOLDOVER mapped to 0, stale when CEP enabled #636

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
37 changes: 37 additions & 0 deletions pkg/daemon/daemon_internal_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,8 @@ import (
"github.com/k8snetworkplumbingwg/linuxptp-daemon/pkg/leap"
ptpv1 "github.com/k8snetworkplumbingwg/ptp-operator/api/v1"
ptpv2alpha1 "github.com/k8snetworkplumbingwg/ptp-operator/api/v2alpha1"
"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/testutil"
"github.com/stretchr/testify/assert"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/client-go/kubernetes/fake"
Expand All @@ -34,6 +36,10 @@ import (
const (
testPtp4lOffsetLine = "ptp4l[1.000]: [ptp4l.0.config] master offset 5 s2 freq -1000 path delay 100"
testSkipStartupReason = "delayed"
testDUTLeadingIface = "ens2f0"
labelIface = "iface"
labelNode = "node"
labelProcess = "process"
)

// vendor defaults are embedded; no filesystem setup needed
Expand Down Expand Up @@ -3546,3 +3552,34 @@ func TestFindProcessesByName(t *testing.T) {
procs = pm.findProcessesByName("nonexistent")
assert.Equal(t, 0, len(procs))
}

func TestUpdateClockStateMetrics(t *testing.T) {
origNodeName := NodeName
defer func() { NodeName = origNodeName }()

NodeName = "test-node"
RegisterMetrics(NodeName)
ClockState.Reset()

tests := []struct {
state string
expected float64
}{
{FREERUN, event.ClockStateFreerun},
{LOCKED, event.ClockStateLocked},
{HOLDOVER, event.ClockStateHoldover},
}

for _, tt := range tests {
t.Run(tt.state, func(t *testing.T) {
updateClockStateMetrics(ptp4lProcessName, testDUTLeadingIface, tt.state)
gauge, err := ClockState.GetMetricWith(prometheus.Labels{
labelProcess: ptp4lProcessName, labelNode: NodeName, labelIface: testDUTLeadingIface,
})
assert.NoError(t, err)
actual := testutil.ToFloat64(gauge)
assert.Equal(t, tt.expected, actual,
"clock_state for %s should be %v", tt.state, tt.expected)
})
}
}
23 changes: 12 additions & 11 deletions pkg/daemon/metrics.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ import (
"time"

"github.com/k8snetworkplumbingwg/linuxptp-daemon/pkg/alias"
"github.com/k8snetworkplumbingwg/linuxptp-daemon/pkg/event"
"github.com/k8snetworkplumbingwg/linuxptp-daemon/pkg/synce"

"github.com/golang/glog"
Expand Down Expand Up @@ -48,12 +49,10 @@ const (
sys = "sys"
)

// Clock state string values used in log parsing and state transitions.
const (
//LOCKED ...
LOCKED string = "LOCKED"
//FREERUN ...
FREERUN = "FREERUN"
// HOLDOVER
LOCKED = "LOCKED"
FREERUN = "FREERUN"
HOLDOVER = "HOLDOVER"
)

Expand Down Expand Up @@ -529,12 +528,14 @@ func updateClockStateMetrics(process, iface string, state string) {
return
}
glog.V(14).Infof("updateClockStateMetrics: process=%s iface=%s state=%s", process, iface, state)
if state == LOCKED {
ClockState.With(prometheus.Labels{
"process": process, "node": NodeName, "iface": iface}).Set(1)
} else {
ClockState.With(prometheus.Labels{
"process": process, "node": NodeName, "iface": iface}).Set(0)
labels := prometheus.Labels{"process": process, "node": NodeName, "iface": iface}
switch state {
case LOCKED:
ClockState.With(labels).Set(event.ClockStateLocked)
case HOLDOVER:
ClockState.With(labels).Set(event.ClockStateHoldover)
default:
ClockState.With(labels).Set(event.ClockStateFreerun)
}
}

Expand Down
25 changes: 14 additions & 11 deletions pkg/event/event.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -150,6 +150,13 @@ const (
PTP_NOTSET PTPState = "-2"
)

// Clock state metric values for openshift_ptp_clock_state gauge.
const (
ClockStateFreerun float64 = 0
ClockStateLocked float64 = 1
ClockStateHoldover float64 = 2
)

const (
// socketDialTimeout is the maximum time to wait for a single dial attempt to the event socket.
socketDialTimeout = 5 * time.Second
Expand Down Expand Up @@ -1129,19 +1136,15 @@ func (e *EventHandler) UpdateClockStateMetrics(state PTPState, process, iFace st
if !utils.CheckMetricSanity("ClockState", process, iFace) {
return
}
if e.stdoutToSocket {
return
}
labels := prometheus.Labels{
"process": process, "node": e.nodeName, "iface": iFace}
if state == PTP_LOCKED {
e.clockMetric.With(labels).Set(1)
} else if state == PTP_FREERUN {
e.clockMetric.With(labels).Set(0)
} else if state == PTP_HOLDOVER {
e.clockMetric.With(labels).Set(2)
} else {
e.clockMetric.With(labels).Set(3)
switch state {
case PTP_LOCKED:
e.clockMetric.With(labels).Set(ClockStateLocked)
case PTP_HOLDOVER:
e.clockMetric.With(labels).Set(ClockStateHoldover)
default:
e.clockMetric.With(labels).Set(ClockStateFreerun)
}
}

Expand Down