IntenseWebs/grafana
3
0
Fork 0
mirror of https://github.com/grafana/grafana.git synced 2025-02-25 18:55:37 -06:00
Code Issues Packages Projects Releases Wiki Activity

Alerting: Add metrics to the remote Alertmanager struct (#79835)

Browse Source 
* Alerting: Add metrics to the remote Alertmanager struct

* rephrase http_requests_failed description

* make linter happy

* remove unnecessary metrics

* extract timed client to separate package

* use histogram collector from dskit

* remove weaveworks dependency

* capture metrics for all requests to the remote Alertmanager (both clients)

* use the timed client in the MimirAuthRoundTripper

* HTTPRequestsDuration -> HTTPRequestDuration, clean up mimir client factory function

* refactor

* less git diff

* gauge for last readiness check in seconds

* initialize LastReadinesCheck to 0, tweak metric names and descriptions

* add counters for sync attempts/errors

* last config sync and last state sync timestamps (gauges)

* change latency metric name

* metric for remote Alertmanager mode

* code review comments

* move label constants to metrics package
This commit is contained in:
Santiago
2024-01-10 11:18:24 +01:00
committed by GitHub
parent 1162c28a55
commit 9e78faa7ba
14 changed files with 171 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
pkg/services/ngalert/remote/alertmanager_test.go
View File

@@ -14,12 +14,14 @@ import (
"github.com/go-openapi/strfmt"
apimodels "github.com/grafana/grafana/pkg/services/ngalert/api/tooling/definitions"
"github.com/grafana/grafana/pkg/services/ngalert/metrics"
ngmodels "github.com/grafana/grafana/pkg/services/ngalert/models"
"github.com/grafana/grafana/pkg/services/ngalert/notifier"
"github.com/grafana/grafana/pkg/services/ngalert/tests/fakes"
"github.com/grafana/grafana/pkg/util"
amv2 "github.com/prometheus/alertmanager/api/v2/models"
"github.com/prometheus/alertmanager/cluster/clusterpb"
"github.com/prometheus/client_golang/prometheus"
"github.com/stretchr/testify/require"
)
@@ -68,7 +70,8 @@ func TestNewAlertmanager(t *testing.T) {
TenantID: test.tenantID,
BasicAuthPassword: test.password,
}
am, err := NewAlertmanager(cfg, nil)
m := metrics.NewRemoteAlertmanagerMetrics(prometheus.NewRegistry())
am, err := NewAlertmanager(cfg, nil, m)
if test.expErr != "" {
require.EqualError(tt, err, test.expErr)
return
@@ -106,7 +109,8 @@ func TestApplyConfig(t *testing.T) {
require.NoError(t, store.Set(ctx, cfg.OrgID, "alertmanager", notifier.SilencesFilename, "test"))
require.NoError(t, store.Set(ctx, cfg.OrgID, "alertmanager", notifier.NotificationLogFilename, "test"))
am, err := NewAlertmanager(cfg, fstore)
m := metrics.NewRemoteAlertmanagerMetrics(prometheus.NewRegistry())
am, err := NewAlertmanager(cfg, fstore, m)
require.NoError(t, err)
config := &ngmodels.AlertConfiguration{}
@@ -175,7 +179,8 @@ func TestIntegrationRemoteAlertmanagerApplyConfigOnlyUploadsOnce(t *testing.T) {
require.NoError(t, err)
encodedFullState := base64.StdEncoding.EncodeToString(fullState)
am, err := NewAlertmanager(cfg, fstore)
m := metrics.NewRemoteAlertmanagerMetrics(prometheus.NewRegistry())
am, err := NewAlertmanager(cfg, fstore, m)
require.NoError(t, err)
// We should have no configuration or state at first.
@@ -259,7 +264,8 @@ func TestIntegrationRemoteAlertmanagerSilences(t *testing.T) {
TenantID: tenantID,
BasicAuthPassword: password,
}
am, err := NewAlertmanager(cfg, nil)
m := metrics.NewRemoteAlertmanagerMetrics(prometheus.NewRegistry())
am, err := NewAlertmanager(cfg, nil, m)
require.NoError(t, err)
// We should have no silences at first.
@@ -339,7 +345,8 @@ func TestIntegrationRemoteAlertmanagerAlerts(t *testing.T) {
TenantID: tenantID,
BasicAuthPassword: password,
}
am, err := NewAlertmanager(cfg, nil)
m := metrics.NewRemoteAlertmanagerMetrics(prometheus.NewRegistry())
am, err := NewAlertmanager(cfg, nil, m)
require.NoError(t, err)
// Wait until the Alertmanager is ready to send alerts.
@@ -405,7 +412,8 @@ func TestIntegrationRemoteAlertmanagerReceivers(t *testing.T) {
BasicAuthPassword: password,
}
am, err := NewAlertmanager(cfg, nil)
m := metrics.NewRemoteAlertmanagerMetrics(prometheus.NewRegistry())
am, err := NewAlertmanager(cfg, nil, m)
require.NoError(t, err)
// We should start with the default config.