Files
zooqueenandHanzo Dev 62b4ca36dc Rule alerts go to the rule's own org
SendAlerts addressed every alert to `SELECT id FROM organizations LIMIT 1`,
which is whichever org sorts first, not the rule's. That org's
alertmanager has none of the rule's channels or route policies, so the
alert sat on its default receiver and reached nobody, and that org's
members could see it. Alerts now carry the rule's own org id.

Co-authored-by: Hanzo Dev <[email protected]>
2026-09-24 16:40:51 -07:00

275 lines
10 KiB
Go

package rules
import (
"context"
"encoding/json"
"math"
"strconv"
"testing"
"time"
"github.com/hanzoai/o11y/pkg/alertmanager"
"github.com/hanzoai/o11y/pkg/alertmanager/alertmanagerserver"
alertmanagermock "github.com/hanzoai/o11y/pkg/alertmanager/alertmanagertest"
"github.com/hanzoai/o11y/pkg/instrumentation/instrumentationtest"
"github.com/hanzoai/o11y/pkg/prometheus"
"github.com/hanzoai/o11y/pkg/prometheus/prometheustest"
"github.com/hanzoai/o11y/pkg/telemetrystore"
"github.com/hanzoai/o11y/pkg/telemetrystore/telemetrystoretest"
"github.com/hanzoai/o11y/pkg/types/alertmanagertypes"
"github.com/hanzoai/o11y/pkg/types/metrictypes"
"github.com/hanzoai/o11y/pkg/valuer"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/mock"
"github.com/stretchr/testify/require"
dsmock "github.com/hanzo-ds/mock"
)
func TestManager_TestNotification_SendUnmatched_ThresholdRule(t *testing.T) {
target := 10.0
recovery := 5.0
for _, tc := range TcTestNotiSendUnmatchedThresholdRule {
t.Run(tc.Name, func(t *testing.T) {
rule := ThresholdRuleAtLeastOnceValueAbove(target, &recovery)
// Marshal rule to JSON as TestNotification expects
ruleBytes, err := json.Marshal(rule)
require.NoError(t, err)
orgID := valuer.GenerateUUID()
// for saving temp alerts that are triggered via TestNotification
triggeredTestAlerts := []map[*alertmanagertypes.PostableAlert][]string{}
// Create manager using test factory with hooks
mgr := NewTestManager(t, &TestManagerOptions{
AlertmanagerHook: func(am alertmanager.Alertmanager) {
mockAM := am.(*alertmanagermock.MockAlertmanager)
// mock set notification config
mockAM.On("SetNotificationConfig", mock.Anything, mock.Anything, mock.Anything, mock.Anything).Return(nil)
mockAM.On("Config").Return(alertmanagerserver.Config{ExternalURL: mustParseURL(t, "http://localhost:8080")})
// for saving temp alerts that are triggered via TestNotification
if tc.ExpectAlerts > 0 {
mockAM.On("TestAlert", mock.Anything, mock.Anything, mock.Anything, mock.Anything).Run(func(args mock.Arguments) {
triggeredTestAlerts = append(triggeredTestAlerts, args.Get(3).(map[*alertmanagertypes.PostableAlert][]string))
}).Return(nil).Times(tc.ExpectAlerts)
}
},
TelemetryStoreHook: func(store telemetrystore.TelemetryStore) {
mockStore := store.(*telemetrystoretest.Provider)
// Set up mock data for telemetry store
cols := make([]dsmock.ColumnType, 0)
cols = append(cols, dsmock.ColumnType{Name: "value", Type: "Float64"})
cols = append(cols, dsmock.ColumnType{Name: "attr", Type: "String"})
cols = append(cols, dsmock.ColumnType{Name: "ts", Type: "DateTime"})
alertDataRows := dsmock.NewRows(cols, tc.Values)
mock := mockStore.Mock()
// Mock metadata queries for FetchTemporalityAndTypeMulti
// First query: fetchMetricsTemporalityAndType (from o11y_metrics time series table)
metadataCols := []dsmock.ColumnType{
{Name: "metric_name", Type: "String"},
{Name: "temporality", Type: "String"},
{Name: "type", Type: "String"},
{Name: "is_monotonic", Type: "Bool"},
}
metadataRows := dsmock.NewRows(metadataCols, [][]any{
{"probe_success", metrictypes.Unspecified, metrictypes.GaugeType, false},
})
mock.ExpectQuery("*distributed_time_series_v4*").WithArgs(nil, nil, nil).WillReturnRows(metadataRows)
// Second query: fetchMeterSourceMetricsTemporalityAndType (from o11y_meter table)
emptyMetadataRows := dsmock.NewRows(metadataCols, [][]any{})
mock.ExpectQuery("*meter*").WithArgs(nil).WillReturnRows(emptyMetadataRows)
// Generate query arguments for the metric query
evalTime := time.Now().UTC()
evalWindow := 5 * time.Minute
evalDelay := time.Duration(0)
queryArgs := GenerateMetricQueryCHArgs(
evalTime,
evalWindow,
evalDelay,
"probe_success",
metrictypes.Unspecified,
)
mock.ExpectQuery("*WITH __temporal_aggregation_cte*").
WithArgs(queryArgs...).
WillReturnRows(alertDataRows)
},
})
count, err := mgr.TestNotification(context.Background(), orgID, string(ruleBytes))
require.Nil(t, err)
assert.Equal(t, tc.ExpectAlerts, count)
if tc.ExpectAlerts > 0 {
// check if the alert has been triggered
require.Len(t, triggeredTestAlerts, 1)
var gotAlerts []*alertmanagertypes.PostableAlert
for a := range triggeredTestAlerts[0] {
gotAlerts = append(gotAlerts, a)
}
require.Len(t, gotAlerts, tc.ExpectAlerts)
// check if the alert has triggered with correct threshold value
if tc.ExpectValue != 0 {
assert.Equal(t, strconv.FormatFloat(tc.ExpectValue, 'f', -1, 64), gotAlerts[0].Annotations["value"])
}
} else {
// check if no alerts have been triggered
assert.Empty(t, triggeredTestAlerts)
}
})
}
}
func TestManager_TestNotification_SendUnmatched_PromRule(t *testing.T) {
target := 10.0
for _, tc := range TcTestNotificationSendUnmatchedPromRule {
t.Run(tc.Name, func(t *testing.T) {
// Capture base time once per test case to ensure consistent timestamps
baseTime := time.Now().UTC()
rule := BuildPromAtLeastOnceValueAbove(target, nil)
// Marshal rule to JSON as TestNotification expects
ruleBytes, err := json.Marshal(rule)
require.NoError(t, err)
orgID := valuer.GenerateUUID()
// for saving temp alerts that are triggered via TestNotification
triggeredTestAlerts := []map[*alertmanagertypes.PostableAlert][]string{}
// Variable to store promProvider for cleanup
var promProvider *prometheustest.Provider
// Create manager using test factory with hooks
mgr := NewTestManager(t, &TestManagerOptions{
AlertmanagerHook: func(am alertmanager.Alertmanager) {
mockAM := am.(*alertmanagermock.MockAlertmanager)
// mock set notification config
mockAM.On("SetNotificationConfig", mock.Anything, mock.Anything, mock.Anything, mock.Anything).Return(nil)
mockAM.On("Config").Return(alertmanagerserver.Config{ExternalURL: mustParseURL(t, "http://localhost:8080")})
// for saving temp alerts that are triggered via TestNotification
if tc.ExpectAlerts > 0 {
mockAM.On("TestAlert", mock.Anything, mock.Anything, mock.Anything, mock.Anything).Run(func(args mock.Arguments) {
triggeredTestAlerts = append(triggeredTestAlerts, args.Get(3).(map[*alertmanagertypes.PostableAlert][]string))
}).Return(nil).Times(tc.ExpectAlerts)
}
},
TelemetryStoreHook: func(store telemetrystore.TelemetryStore) {
mockStore := store.(*telemetrystoretest.Provider)
// Set up Prometheus-specific mock data
// Fingerprint columns for Prometheus queries
fingerprintCols := []dsmock.ColumnType{
{Name: "fingerprint", Type: "UInt64"},
{Name: "any(labels)", Type: "String"},
}
// Samples columns for Prometheus queries
samplesCols := []dsmock.ColumnType{
{Name: "metric_name", Type: "String"},
{Name: "fingerprint", Type: "UInt64"},
{Name: "unix_milli", Type: "Int64"},
{Name: "value", Type: "Float64"},
{Name: "flags", Type: "UInt32"},
}
// Calculate query time range similar to Prometheus rule tests
// TestNotification uses time.Now().UTC() for evaluation
// We calculate the query window based on current time to match what the actual evaluation will use
evalTime := baseTime
evalWindowMs := int64(5 * 60 * 1000) // 5 minutes in ms
evalTimeMs := evalTime.UnixMilli()
queryStart := ((evalTimeMs-2*evalWindowMs)/60000)*60000 + 1 // truncate to minute + 1ms
queryEnd := (evalTimeMs / 60000) * 60000 // truncate to minute
// Create fingerprint data
fingerprint := uint64(12345)
labelsJSON := `{"__name__":"test_metric"}`
fingerprintData := [][]any{
{fingerprint, labelsJSON},
}
fingerprintRows := dsmock.NewRows(fingerprintCols, fingerprintData)
// Create samples data from test case values, calculating timestamps relative to baseTime
validSamplesData := make([][]any, 0)
for _, v := range tc.Values {
// Skip NaN and Inf values in the samples data
if math.IsNaN(v.Value) || math.IsInf(v.Value, 0) {
continue
}
// Calculate timestamp relative to baseTime
sampleTimestamp := baseTime.Add(v.Offset).UnixMilli()
validSamplesData = append(validSamplesData, []any{
"test_metric",
fingerprint,
sampleTimestamp,
v.Value,
uint32(0), // flags - 0 means normal value
})
}
samplesRows := dsmock.NewRows(samplesCols, validSamplesData)
mock := mockStore.Mock()
// Mock the fingerprint query (for Prometheus label matching)
mock.ExpectQuery("SELECT fingerprint, any").
WithArgs("test_metric", "__name__", "test_metric").
WillReturnRows(fingerprintRows)
// Mock the samples query (for Prometheus metric data)
mock.ExpectQuery("SELECT metric_name, fingerprint, unix_milli").
WithArgs(
"test_metric",
"test_metric",
"__name__",
"test_metric",
queryStart,
queryEnd,
).
WillReturnRows(samplesRows)
// Create Prometheus provider for this test
promProvider = prometheustest.New(context.Background(), instrumentationtest.New().ToProviderSettings(), prometheus.Config{Timeout: 2 * time.Minute}, store)
},
ManagerOptionsHook: func(opts *ManagerOptions) {
// Set Prometheus provider for PromQL queries
if promProvider != nil {
opts.Prometheus = promProvider
}
},
})
count, err := mgr.TestNotification(context.Background(), orgID, string(ruleBytes))
require.Nil(t, err)
assert.Equal(t, tc.ExpectAlerts, count)
if tc.ExpectAlerts > 0 {
// check if the alert has been triggered
require.Len(t, triggeredTestAlerts, 1)
var gotAlerts []*alertmanagertypes.PostableAlert
for a := range triggeredTestAlerts[0] {
gotAlerts = append(gotAlerts, a)
}
require.Len(t, gotAlerts, tc.ExpectAlerts)
// check if the alert has triggered with correct threshold value
if tc.ExpectValue != 0 && !math.IsNaN(tc.ExpectValue) && !math.IsInf(tc.ExpectValue, 0) {
assert.Equal(t, strconv.FormatFloat(tc.ExpectValue, 'f', -1, 64), gotAlerts[0].Annotations["value"])
}
} else {
// check if no alerts have been triggered
assert.Empty(t, triggeredTestAlerts)
}
promProvider.Close()
})
}
}