mirror of
https://github.com/grafana/grafana.git
synced 2024-11-25 18:30:41 -06:00
f6a46744a6
Backend: * Update the Grafana Alerting engine to provide feedback to HysteresisCommand. The feedback information is stored in state.Manager as a fingerprint of each state. The fingerprint is persisted to the database. Only fingerprints that belong to Pending and Alerting states are considered as "loaded" and provided back to the command. - add ResultFingerprint to state.State. It's different from other fingerprints we store in the state because it is calculated from the result labels. - add rule_fingerprint column to alert_instance - update alerting evaluator to accept AlertingResultsReader via context, and update scheduler to provide it. - add AlertingResultsFromRuleState that implements the new interface in eval package - update getExprRequest to patch the hysteresis command. * Only one "Recovery Threshold" query is allowed to be used in the alert rule and it must be the Condition. Frontend: * Add hysteresis option to Threshold in UI. It's called "Recovery Threshold" * Add test for getUnloadEvaluatorTypeFromCondition * Hide hysteresis in panel expressions * Refactor isInvalid and add test for it * Remove unnecesary React.memo * Add tests for updateEvaluatorConditions --------- Co-authored-by: Sonia Aguilar <soniaaguilarpeiron@gmail.com>
410 lines
12 KiB
Go
410 lines
12 KiB
Go
package backtesting
|
|
|
|
import (
|
|
"context"
|
|
"encoding/json"
|
|
"errors"
|
|
"fmt"
|
|
"math/rand"
|
|
"testing"
|
|
"time"
|
|
|
|
"github.com/stretchr/testify/require"
|
|
|
|
"github.com/grafana/grafana-plugin-sdk-go/data"
|
|
|
|
"github.com/grafana/grafana/pkg/services/auth/identity"
|
|
"github.com/grafana/grafana/pkg/services/ngalert/eval"
|
|
"github.com/grafana/grafana/pkg/services/ngalert/eval/eval_mocks"
|
|
"github.com/grafana/grafana/pkg/services/ngalert/models"
|
|
"github.com/grafana/grafana/pkg/services/ngalert/state"
|
|
"github.com/grafana/grafana/pkg/util"
|
|
)
|
|
|
|
func TestNewBacktestingEvaluator(t *testing.T) {
|
|
t.Run("creates data evaluator", func(t *testing.T) {
|
|
frame := GenerateWideSeriesFrame(10, time.Second)
|
|
d := struct {
|
|
Data *data.Frame `json:"data"`
|
|
}{
|
|
Data: frame,
|
|
}
|
|
validData, err := json.Marshal(d)
|
|
require.NoError(t, err)
|
|
refID := util.GenerateShortUID()
|
|
|
|
evalFactory := eval_mocks.NewEvaluatorFactory(&eval_mocks.ConditionEvaluatorMock{})
|
|
|
|
testCases := []struct {
|
|
name string
|
|
condition models.Condition
|
|
error bool
|
|
expectedEval backtestingEvaluator
|
|
}{
|
|
{
|
|
name: "creates data evaluator when there is one query with type __data__",
|
|
condition: models.Condition{
|
|
Condition: refID,
|
|
Data: []models.AlertQuery{
|
|
{
|
|
RefID: refID,
|
|
QueryType: "__data__",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: "",
|
|
Model: json.RawMessage(validData),
|
|
},
|
|
},
|
|
},
|
|
expectedEval: &dataEvaluator{},
|
|
},
|
|
{
|
|
name: "creates data evaluator when there is one query with datasource UID __data__",
|
|
condition: models.Condition{
|
|
Condition: refID,
|
|
Data: []models.AlertQuery{
|
|
{
|
|
RefID: refID,
|
|
QueryType: "",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: "__data__",
|
|
Model: json.RawMessage(validData),
|
|
},
|
|
},
|
|
},
|
|
expectedEval: &dataEvaluator{},
|
|
}, {
|
|
name: "fails if queries contain data and other queries",
|
|
condition: models.Condition{
|
|
Condition: refID,
|
|
Data: []models.AlertQuery{
|
|
{
|
|
RefID: refID,
|
|
QueryType: "__data__",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: "",
|
|
Model: json.RawMessage(validData),
|
|
},
|
|
{
|
|
RefID: "D",
|
|
QueryType: "",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: util.GenerateShortUID(),
|
|
},
|
|
},
|
|
},
|
|
error: true,
|
|
},
|
|
{
|
|
name: "fails if data query does not contain data",
|
|
condition: models.Condition{
|
|
Condition: refID,
|
|
Data: []models.AlertQuery{
|
|
{
|
|
RefID: refID,
|
|
QueryType: "__data__",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: "",
|
|
Model: json.RawMessage(nil),
|
|
},
|
|
},
|
|
},
|
|
error: true,
|
|
},
|
|
{
|
|
name: "fails if data query does not contain frame in data",
|
|
condition: models.Condition{
|
|
Condition: refID,
|
|
Data: []models.AlertQuery{
|
|
{
|
|
RefID: refID,
|
|
QueryType: "__data__",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: "",
|
|
Model: json.RawMessage(`{ "data": "test"}`),
|
|
},
|
|
},
|
|
},
|
|
error: true,
|
|
}, {
|
|
name: "fails if condition refID and data refID does not match",
|
|
condition: models.Condition{
|
|
Condition: refID,
|
|
Data: []models.AlertQuery{
|
|
{
|
|
RefID: "B",
|
|
QueryType: "__data__",
|
|
RelativeTimeRange: models.RelativeTimeRange{},
|
|
DatasourceUID: "",
|
|
Model: json.RawMessage(validData),
|
|
},
|
|
},
|
|
},
|
|
error: true,
|
|
},
|
|
}
|
|
|
|
for _, testCase := range testCases {
|
|
t.Run(testCase.name, func(t *testing.T) {
|
|
e, err := newBacktestingEvaluator(context.Background(), evalFactory, nil, testCase.condition, nil)
|
|
if testCase.error {
|
|
require.Error(t, err)
|
|
return
|
|
}
|
|
require.NoError(t, err)
|
|
require.IsType(t, &dataEvaluator{}, e)
|
|
})
|
|
}
|
|
})
|
|
}
|
|
|
|
func TestEvaluatorTest(t *testing.T) {
|
|
states := []eval.State{eval.Normal, eval.Alerting, eval.Pending}
|
|
generateState := func(prefix string) *state.State {
|
|
return &state.State{
|
|
CacheID: "state-" + prefix,
|
|
Labels: models.GenerateAlertLabels(rand.Intn(5)+1, prefix+"-"),
|
|
State: states[rand.Intn(len(states))],
|
|
}
|
|
}
|
|
|
|
randomResultCallback := func(now time.Time) (eval.Results, error) {
|
|
return eval.GenerateResults(rand.Intn(5)+1, eval.ResultGen()), nil
|
|
}
|
|
evaluator := &fakeBacktestingEvaluator{
|
|
evalCallback: randomResultCallback,
|
|
}
|
|
manager := &fakeStateManager{}
|
|
|
|
backtestingEvaluatorFactory = func(ctx context.Context, evalFactory eval.EvaluatorFactory, user identity.Requester, condition models.Condition, r eval.AlertingResultsReader) (backtestingEvaluator, error) {
|
|
return evaluator, nil
|
|
}
|
|
|
|
t.Cleanup(func() {
|
|
backtestingEvaluatorFactory = newBacktestingEvaluator
|
|
})
|
|
|
|
engine := &Engine{
|
|
evalFactory: nil,
|
|
createStateManager: func() stateManager {
|
|
return manager
|
|
},
|
|
}
|
|
rule := models.AlertRuleGen(models.WithInterval(time.Second))()
|
|
ruleInterval := time.Duration(rule.IntervalSeconds) * time.Second
|
|
|
|
t.Run("should return data frame in specific format", func(t *testing.T) {
|
|
from := time.Unix(0, 0)
|
|
to := from.Add(5 * ruleInterval)
|
|
allStates := [...]eval.State{eval.Normal, eval.Alerting, eval.Pending, eval.NoData, eval.Error}
|
|
|
|
var states []state.StateTransition
|
|
|
|
for _, s := range allStates {
|
|
states = append(states, state.StateTransition{
|
|
State: &state.State{
|
|
CacheID: "state-" + s.String(),
|
|
Labels: models.GenerateAlertLabels(rand.Intn(5)+1, s.String()+"-"),
|
|
State: s,
|
|
StateReason: util.GenerateShortUID(),
|
|
},
|
|
})
|
|
}
|
|
|
|
manager.stateCallback = func(now time.Time) []state.StateTransition {
|
|
return states
|
|
}
|
|
|
|
frame, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
|
|
require.NoError(t, err)
|
|
require.Len(t, frame.Fields, len(states)+1) // +1 - timestamp
|
|
|
|
t.Run("should contain field Time", func(t *testing.T) {
|
|
timestampField, _ := frame.FieldByName("Time")
|
|
require.NotNil(t, timestampField, "frame does not contain field 'Time'")
|
|
require.Equal(t, data.FieldTypeTime, timestampField.Type())
|
|
})
|
|
|
|
fieldByState := make(map[string]*data.Field, len(states))
|
|
|
|
t.Run("should contain a field per state", func(t *testing.T) {
|
|
for _, s := range states {
|
|
var f *data.Field
|
|
for _, field := range frame.Fields {
|
|
if field.Labels.String() == s.Labels.String() {
|
|
f = field
|
|
break
|
|
}
|
|
}
|
|
require.NotNilf(t, f, "Cannot find a field by state labels")
|
|
fieldByState[s.CacheID] = f
|
|
}
|
|
})
|
|
|
|
t.Run("should be populated with correct values", func(t *testing.T) {
|
|
timestampField, _ := frame.FieldByName("Time")
|
|
expectedLength := timestampField.Len()
|
|
for _, field := range frame.Fields {
|
|
require.Equalf(t, expectedLength, field.Len(), "Field %s should have the size %d", field.Name, expectedLength)
|
|
}
|
|
for i := 0; i < expectedLength; i++ {
|
|
expectedTime := from.Add(time.Duration(int64(i)*rule.IntervalSeconds) * time.Second)
|
|
require.Equal(t, expectedTime, timestampField.At(i).(time.Time))
|
|
for _, s := range states {
|
|
f := fieldByState[s.CacheID]
|
|
if s.State.State == eval.NoData {
|
|
require.Nil(t, f.At(i))
|
|
} else {
|
|
v := f.At(i).(*string)
|
|
require.NotNilf(t, v, "Field [%s] value at index %d should not be nil", s.CacheID, i)
|
|
require.Equal(t, fmt.Sprintf("%s (%s)", s.State.State, s.StateReason), *v)
|
|
}
|
|
}
|
|
}
|
|
})
|
|
})
|
|
|
|
t.Run("should not fail if 'to-from' is not times of interval", func(t *testing.T) {
|
|
from := time.Unix(0, 0)
|
|
to := from.Add(5 * ruleInterval)
|
|
|
|
states := []state.StateTransition{
|
|
{
|
|
State: &state.State{
|
|
CacheID: "state-1",
|
|
Labels: models.GenerateAlertLabels(rand.Intn(5)+1, "test-"),
|
|
State: eval.Normal,
|
|
StateReason: util.GenerateShortUID(),
|
|
},
|
|
},
|
|
}
|
|
|
|
manager.stateCallback = func(now time.Time) []state.StateTransition {
|
|
return states
|
|
}
|
|
|
|
frame, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
require.NoError(t, err)
|
|
expectedLen := frame.Rows()
|
|
for i := 0; i < 100; i++ {
|
|
jitter := time.Duration(rand.Int63n(ruleInterval.Milliseconds())) * time.Millisecond
|
|
frame, err = engine.Test(context.Background(), nil, rule, from, to.Add(jitter))
|
|
require.NoError(t, err)
|
|
require.Equalf(t, expectedLen, frame.Rows(), "jitter %v caused result to be different that base-line", jitter)
|
|
}
|
|
})
|
|
|
|
t.Run("should backfill field with nulls if a new dimension created in the middle", func(t *testing.T) {
|
|
from := time.Unix(0, 0)
|
|
|
|
state1 := state.StateTransition{
|
|
State: generateState("1"),
|
|
}
|
|
state2 := state.StateTransition{
|
|
State: generateState("2"),
|
|
}
|
|
state3 := state.StateTransition{
|
|
State: generateState("3"),
|
|
}
|
|
stateByTime := map[time.Time][]state.StateTransition{
|
|
from: {state1, state2},
|
|
from.Add(1 * ruleInterval): {state1, state2},
|
|
from.Add(2 * ruleInterval): {state1, state2},
|
|
from.Add(3 * ruleInterval): {state1, state2, state3},
|
|
from.Add(4 * ruleInterval): {state1, state2, state3},
|
|
}
|
|
to := from.Add(time.Duration(len(stateByTime)) * ruleInterval)
|
|
|
|
manager.stateCallback = func(now time.Time) []state.StateTransition {
|
|
return stateByTime[now]
|
|
}
|
|
|
|
frame, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
require.NoError(t, err)
|
|
|
|
var field3 *data.Field
|
|
for _, field := range frame.Fields {
|
|
if field.Labels.String() == state3.Labels.String() {
|
|
field3 = field
|
|
break
|
|
}
|
|
}
|
|
require.NotNilf(t, field3, "Result for state 3 was not found")
|
|
require.Equalf(t, len(stateByTime), field3.Len(), "State3 result has unexpected number of values")
|
|
|
|
idx := 0
|
|
for curTime, states := range stateByTime {
|
|
value := field3.At(idx).(*string)
|
|
if len(states) == 2 {
|
|
require.Nilf(t, value, "The result should be nil if state3 was not available for time %v", curTime)
|
|
}
|
|
}
|
|
})
|
|
|
|
t.Run("should fail", func(t *testing.T) {
|
|
manager.stateCallback = func(now time.Time) []state.StateTransition {
|
|
return nil
|
|
}
|
|
|
|
t.Run("when interval is not correct", func(t *testing.T) {
|
|
from := time.Now()
|
|
t.Run("when from=to", func(t *testing.T) {
|
|
to := from
|
|
_, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
require.ErrorIs(t, err, ErrInvalidInputData)
|
|
})
|
|
t.Run("when from > to", func(t *testing.T) {
|
|
to := from.Add(-ruleInterval)
|
|
_, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
require.ErrorIs(t, err, ErrInvalidInputData)
|
|
})
|
|
t.Run("when to-from < interval", func(t *testing.T) {
|
|
to := from.Add(ruleInterval).Add(-time.Millisecond)
|
|
_, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
require.ErrorIs(t, err, ErrInvalidInputData)
|
|
})
|
|
})
|
|
|
|
t.Run("when evaluation fails", func(t *testing.T) {
|
|
expectedError := errors.New("test-error")
|
|
evaluator.evalCallback = func(now time.Time) (eval.Results, error) {
|
|
return nil, expectedError
|
|
}
|
|
from := time.Now()
|
|
to := from.Add(ruleInterval)
|
|
_, err := engine.Test(context.Background(), nil, rule, from, to)
|
|
require.ErrorIs(t, err, expectedError)
|
|
})
|
|
})
|
|
}
|
|
|
|
type fakeStateManager struct {
|
|
stateCallback func(now time.Time) []state.StateTransition
|
|
}
|
|
|
|
func (f *fakeStateManager) ProcessEvalResults(_ context.Context, evaluatedAt time.Time, _ *models.AlertRule, _ eval.Results, _ data.Labels) []state.StateTransition {
|
|
return f.stateCallback(evaluatedAt)
|
|
}
|
|
|
|
func (f *fakeStateManager) GetStatesForRuleUID(orgID int64, alertRuleUID string) []*state.State {
|
|
return nil
|
|
}
|
|
|
|
type fakeBacktestingEvaluator struct {
|
|
evalCallback func(now time.Time) (eval.Results, error)
|
|
}
|
|
|
|
func (f *fakeBacktestingEvaluator) Eval(_ context.Context, from time.Time, interval time.Duration, evaluations int, callback callbackFunc) error {
|
|
for idx, now := 0, from; idx < evaluations; idx, now = idx+1, now.Add(interval) {
|
|
results, err := f.evalCallback(now)
|
|
if err != nil {
|
|
return err
|
|
}
|
|
err = callback(idx, now, results)
|
|
if err != nil {
|
|
return err
|
|
}
|
|
}
|
|
return nil
|
|
}
|