Add sdnotify event at startup and stop to be used with systemd type=n…

…otify (#8541)

Co-authored-by: Ying WANG <74549700+ying-jeanne@users.noreply.github.com>

CHANGELOG.md

@@ -26,6 +26,7 @@
 * [ENHANCEMENT] Query-frontend: added `remote_read` to `op` supported label values for the `cortex_query_frontend_queries_total` metric. #8412
 * [ENHANCEMENT] Query-frontend: log the overall length and start, end time offset from current time for remote read requests. The start and end times are calculated as the miminum and maximum times of the individual queries in the remote read request. #8404
 * [ENHANCEMENT] Storage Provider: Added option `-<prefix>.s3.dualstack-enabled` that allows disabling S3 client from resolving AWS S3 endpoint into dual-stack IPv4/IPv6 endpoint. Defaults to true. #8405
+* [ENHANCEMENT] Use sd_notify to send events at start and stop of services. To be used when running with systemd and Type=notify. #8220
 * [ENHANCEMENT] Alertmanager: Reloading config and templates no longer needs to hit the disk. #4967
 * [ENHANCEMENT] Compactor: Added experimantal `-compactor.in-memory-tenant-meta-cache-size` option to set size of in-memory cache (in number of items) for parsed meta.json files. This can help when tenant has many meta.json files and their parsing before each compaction cycle is using a lot of CPU time. #8544
 * [ENHANCEMENT] Distributor: Interrupt OTLP write request translation when context is canceled or has timed out. #8524

go.mod

@@ -69,6 +69,7 @@ require (
 	github.com/hashicorp/vault/api v1.10.0
 	github.com/mitchellh/colorstring v0.0.0-20190213212951-d06e56a500db
 	github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822
+	github.com/okzk/sdnotify v0.0.0-20180710141335-d9becc38acbd
 	github.com/prometheus/procfs v0.15.1
 	github.com/shirou/gopsutil/v4 v4.24.5
 	github.com/thanos-io/objstore v0.0.0-20240617083302-124528d695c2

go.sum

@@ -752,6 +752,8 @@ github.com/oklog/run v1.1.0 h1:GEenZ1cK0+q0+wsJew9qUg/DyD8k3JzYsZAi5gYi2mA=
 github.com/oklog/run v1.1.0/go.mod h1:sVPdnTZT1zYwAJeCMu2Th4T21pA3FPOQRfWjQlk7DVU=
 github.com/oklog/ulid v1.3.1 h1:EGfNDEx6MqHz8B3uNV6QAib1UR2Lm97sHi3ocA6ESJ4=
 github.com/oklog/ulid v1.3.1/go.mod h1:CirwcVhetQ6Lv90oh/F+FBtV6XMibvdAFo93nm5qn4U=
+github.com/okzk/sdnotify v0.0.0-20180710141335-d9becc38acbd h1:+iAPaTbi1gZpcpDwe/BW1fx7Xoesv69hLNGPheoyhBs=
+github.com/okzk/sdnotify v0.0.0-20180710141335-d9becc38acbd/go.mod h1:4soZNh0zW0LtYGdQ416i0jO0EIqMGcbtaspRS4BDvRQ=
 github.com/onsi/ginkgo v1.16.5 h1:8xi0RTUf59SOSfEtZMvwTvXYMzG4gV23XVHOZiXNtnE=
 github.com/onsi/ginkgo v1.16.5/go.mod h1:+E8gABHa3K6zRBolWtd+ROzc/U5bkGt0FwiG042wbpU=
 github.com/onsi/gomega v1.29.0 h1:KIA/t2t5UBzoirT4H9tsML45GEbo3ouUnBHsCfD2tVg=

pkg/mimir/mimir.go

@@ -31,6 +31,7 @@ import (
 	"github.com/grafana/dskit/services"
 	"github.com/grafana/dskit/signals"
 	"github.com/grafana/dskit/spanprofiler"
+	"github.com/okzk/sdnotify"
 	"github.com/opentracing/opentracing-go"
 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
@@ -886,9 +887,15 @@ func (t *Mimir) Run() error {
 		grpcutil.WithManager(sm),
 	))
 
-	// Let's listen for events from this manager, and log them.
-	healthy := func() { level.Info(util_log.Logger).Log("msg", "Application started") }
-	stopped := func() { level.Info(util_log.Logger).Log("msg", "Application stopped") }
+	// Let's listen for events from this manager, log them and send sd_notify event
+	healthy := func() {
+		level.Info(util_log.Logger).Log("msg", "Application started")
+		_ = sdnotify.Ready()
+	}
+	stopped := func() {
+		level.Info(util_log.Logger).Log("msg", "Application stopped")
+		_ = sdnotify.Stopping()
+	}
 	serviceFailed := func(service services.Service) {
 		// if any service fails, stop entire Mimir
 		sm.StopAsync()