vmalert: fix variable $activeAt value when templating rule annotation in replay mode

Co-authored-by: Roman Khavronenko <roman@victoriametrics.com>
This commit is contained in:
Hui Wang 2024-09-20 17:07:40 +08:00 committed by GitHub
parent 6167bccc5a
commit d6d02d7aeb
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 17 additions and 17 deletions

View file

@ -335,6 +335,8 @@ func (ar *AlertingRule) execRange(ctx context.Context, start, end time.Time) ([]
// reset to Pending if there are gaps > EvalInterval between DPs
a.State = notifier.StatePending
a.ActiveAt = at
// re-template the annotations as active timestamp is changed
_, a.Annotations, _ = ar.expandTemplates(s, qFn, at)
a.Start = time.Time{}
} else if at.Sub(a.ActiveAt) >= ar.For && a.State != notifier.StateFiring {
a.State = notifier.StateFiring

View file

@ -289,7 +289,7 @@ func TestAlertingRule_Exec(t *testing.T) {
4: {{labels: []string{"name", "foo"}, alert: &notifier.Alert{State: notifier.StateFiring}}},
})
f(newTestAlertingRuleWithKeepFiring("for-pending=>firing=>keepfiring=>firing", defaultStep, defaultStep), [][]datasource.Metric{
f(newTestAlertingRuleWithCustomFields("for-pending=>firing=>keepfiring=>firing", defaultStep, 0, defaultStep, nil), [][]datasource.Metric{
{metricWithLabels(t, "name", "foo")},
{metricWithLabels(t, "name", "foo")},
// empty step to keep firing
@ -302,7 +302,7 @@ func TestAlertingRule_Exec(t *testing.T) {
3: {{labels: []string{"name", "foo"}, alert: &notifier.Alert{State: notifier.StateFiring}}},
})
f(newTestAlertingRuleWithKeepFiring("for-pending=>firing=>keepfiring=>keepfiring=>inactive=>pending=>firing", defaultStep, 2*defaultStep), [][]datasource.Metric{
f(newTestAlertingRuleWithCustomFields("for-pending=>firing=>keepfiring=>keepfiring=>inactive=>pending=>firing", defaultStep, 0, 2*defaultStep, nil), [][]datasource.Metric{
{metricWithLabels(t, "name", "foo")},
{metricWithLabels(t, "name", "foo")},
// empty step to keep firing
@ -395,7 +395,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
{State: notifier.StateFiring, ActiveAt: time.Unix(3e3, 0)},
}, nil)
f(newTestAlertingRule("for-pending", time.Second), []datasource.Metric{
f(newTestAlertingRuleWithCustomFields("for-pending", time.Second, 0, 0, map[string]string{"activeAt": "{{ $activeAt.UnixMilli }}"}), []datasource.Metric{
{Values: []float64{1, 1, 1}, Timestamps: []int64{1, 3, 5}},
}, []*notifier.Alert{
{State: notifier.StatePending, ActiveAt: time.Unix(1, 0)},
@ -406,7 +406,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
GroupID: fakeGroup.ID(),
Name: "for-pending",
Labels: map[string]string{"alertname": "for-pending"},
Annotations: map[string]string{},
Annotations: map[string]string{"activeAt": "5000"},
State: notifier.StatePending,
ActiveAt: time.Unix(5, 0),
Value: 1,
@ -414,7 +414,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
},
})
f(newTestAlertingRule("for-firing", 3*time.Second), []datasource.Metric{
f(newTestAlertingRuleWithCustomFields("for-firing", 3*time.Second, 0, 0, map[string]string{"activeAt": "{{ $activeAt.UnixMilli }}"}), []datasource.Metric{
{Values: []float64{1, 1, 1}, Timestamps: []int64{1, 3, 5}},
}, []*notifier.Alert{
{State: notifier.StatePending, ActiveAt: time.Unix(1, 0)},
@ -425,7 +425,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
GroupID: fakeGroup.ID(),
Name: "for-firing",
Labels: map[string]string{"alertname": "for-firing"},
Annotations: map[string]string{},
Annotations: map[string]string{"activeAt": "1000"},
State: notifier.StateFiring,
ActiveAt: time.Unix(1, 0),
Start: time.Unix(5, 0),
@ -434,7 +434,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
},
})
f(newTestAlertingRule("for-hold-pending", time.Second), []datasource.Metric{
f(newTestAlertingRuleWithCustomFields("for-hold-pending", time.Second, 0, 0, map[string]string{"activeAt": "{{ $activeAt.UnixMilli }}"}), []datasource.Metric{
{Values: []float64{1, 1, 1}, Timestamps: []int64{1, 2, 5}},
}, []*notifier.Alert{
{State: notifier.StatePending, ActiveAt: time.Unix(1, 0)},
@ -445,7 +445,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
GroupID: fakeGroup.ID(),
Name: "for-hold-pending",
Labels: map[string]string{"alertname": "for-hold-pending"},
Annotations: map[string]string{},
Annotations: map[string]string{"activeAt": "5000"},
State: notifier.StatePending,
ActiveAt: time.Unix(5, 0),
Value: 1,
@ -453,7 +453,7 @@ func TestAlertingRuleExecRange(t *testing.T) {
},
})
f(newTestAlertingRuleWithEvalInterval("firing=>inactive=>inactive=>firing=>firing", 0, time.Second), []datasource.Metric{
f(newTestAlertingRuleWithCustomFields("firing=>inactive=>inactive=>firing=>firing", 0, time.Second, 0, nil), []datasource.Metric{
{Values: []float64{1, 1, 1, 1}, Timestamps: []int64{1, 4, 5, 6}},
}, []*notifier.Alert{
{State: notifier.StateFiring, ActiveAt: time.Unix(1, 0)},
@ -538,7 +538,6 @@ func TestAlertingRuleExecRange(t *testing.T) {
"source": "vm",
},
},
//
{
State: notifier.StateFiring, ActiveAt: time.Unix(1, 0),
Labels: map[string]string{
@ -1036,15 +1035,13 @@ func newTestAlertingRule(name string, waitFor time.Duration) *AlertingRule {
return &rule
}
func newTestAlertingRuleWithEvalInterval(name string, waitFor, evalInterval time.Duration) *AlertingRule {
func newTestAlertingRuleWithCustomFields(name string, waitFor, evalInterval, keepFiringFor time.Duration, annotation map[string]string) *AlertingRule {
rule := newTestAlertingRule(name, waitFor)
if evalInterval != 0 {
rule.EvalInterval = evalInterval
return rule
}
func newTestAlertingRuleWithKeepFiring(name string, waitFor, keepFiringFor time.Duration) *AlertingRule {
rule := newTestAlertingRule(name, waitFor)
rule.KeepFiringFor = keepFiringFor
rule.Annotations = annotation
return rule
}

View file

@ -39,6 +39,7 @@ See also [LTS releases](https://docs.victoriametrics.com/lts-releases/).
* BUGFIX: [Single-node VictoriaMetrics](https://docs.victoriametrics.com/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/cluster-victoriametrics/): properly ingest stale NaN samples. Previously it could be dropped if series didn't exist at storage node. See this [issue](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5069) for details.
* BUGFIX: [Single-node VictoriaMetrics](https://docs.victoriametrics.com/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/cluster-victoriametrics/): properly track `vm_missing_tsids_for_metric_id_total` metric. See this [issue](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/6931) for details.
* BUGFIX: [vmalert](https://docs.victoriametrics.com/vmalert): do not send notifications without labels to Alertmanager. Such notifications are rejected by Alertmanager anyway. Before, vmalert could send alert notifications even if no label-value pairs left after applying `alert_relabel_configs` from [notifier config](https://docs.victoriametrics.com/vmalert/#notifier-configuration-file).
* BUGFIX: [vmalert](https://docs.victoriametrics.com/vmalert/): properly update value of variable `$activeAt` in rules annotation during replay mode. Before, `$activeAt` could have provided incorrect values during replay.
* BUGFIX: [MetricsQL](https://docs.victoriametrics.com/metricsql/): properly handle `c1 AND c2` and `c1 OR c1` queries for constants `c1` and `c2`. Previously such queries could return unexpected results. See [this issue](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/6637).
* BUGFIX: all VictoriaMetrics components: increase default value of `-loggerMaxArgLen` cmd-line flag from 1000 to 5000. This should improve visibility on errors produced by very long queries.