From c4848f9b4fbd365daa3285990d74cb9aa89d0267 Mon Sep 17 00:00:00 2001
From: Ritesh H Shukla <kerneltime@gmail.com>
Date: Mon, 1 Feb 2021 23:02:18 -0800
Subject: [PATCH] Add process start time to cluster metrics. (#11405)

---
 cmd/metrics-v2.go               | 32 +++++++++++++++++++++++++++++---
 docs/metrics/prometheus/list.md |  1 +
 2 files changed, 30 insertions(+), 3 deletions(-)

diff --git a/cmd/metrics-v2.go b/cmd/metrics-v2.go
index b08fc5fb4..027e3f937 100644
--- a/cmd/metrics-v2.go
+++ b/cmd/metrics-v2.go
@@ -53,18 +53,19 @@ const (
 	capacityRawSubsystem    MetricSubsystem = "capacity_raw"
 	capacityUsableSubsystem MetricSubsystem = "capacity_usable"
 	diskSubsystem           MetricSubsystem = "disk"
+	fileDescriptorSubsystem MetricSubsystem = "file_descriptor"
 	goRoutines              MetricSubsystem = "go_routine"
+	ioSubsystem             MetricSubsystem = "io"
 	nodesSubsystem          MetricSubsystem = "nodes"
 	objectsSubsystem        MetricSubsystem = "objects"
-	fileDescriptorSubsystem MetricSubsystem = "file_descriptor"
-	ioSubsystem             MetricSubsystem = "io"
+	processSubsystem        MetricSubsystem = "process"
 	replicationSubsystem    MetricSubsystem = "replication"
 	requestsSubsystem       MetricSubsystem = "requests"
 	timeSubsystem           MetricSubsystem = "time"
 	trafficSubsystem        MetricSubsystem = "traffic"
+	softwareSubsystem       MetricSubsystem = "software"
 	sysCallSubsystem        MetricSubsystem = "syscall"
 	usageSubsystem          MetricSubsystem = "usage"
-	softwareSubsystem       MetricSubsystem = "software"
 )
 
 // MetricName are the individual names for the metric.
@@ -107,6 +108,7 @@ const (
 	ttfbDistribution = "ttbf_seconds_distribution"
 
 	lastActivityTime = "last_activity_nano_seconds"
+	startTime        = "starttime_seconds"
 )
 
 const (
@@ -631,6 +633,15 @@ func getMinIOGORoutineCountMD() MetricDescription {
 		Type:      gaugeMetric,
 	}
 }
+func getMinIOProcessStartTimeMD() MetricDescription {
+	return MetricDescription{
+		Namespace: nodeMetricNamespace,
+		Subsystem: processSubsystem,
+		Name:      startTime,
+		Help:      "Start time for MinIO process per node in seconds.",
+		Type:      gaugeMetric,
+	}
+}
 func getMinioProcMetrics() MetricsGroup {
 	return MetricsGroup{
 		Metrics: []Metric{},
@@ -656,6 +667,16 @@ func getMinioProcMetrics() MetricsGroup {
 				logger.LogOnceIf(ctx, err, ioSubsystem)
 				return
 			}
+			stat, err := p.Stat()
+			if err != nil {
+				logger.LogOnceIf(ctx, err, processSubsystem)
+				return
+			}
+			startTime, err := stat.StartTime()
+			if err != nil {
+				logger.LogOnceIf(ctx, err, startTime)
+				return
+			}
 
 			metrics.Metrics = append(metrics.Metrics,
 				Metric{
@@ -698,6 +719,11 @@ func getMinioProcMetrics() MetricsGroup {
 					Description: getMinioProcessIOWriteCachedBytesMD(),
 					Value:       float64(io.WChar),
 				})
+			metrics.Metrics = append(metrics.Metrics,
+				Metric{
+					Description: getMinIOProcessStartTimeMD(),
+					Value:       startTime,
+				})
 		},
 	}
 }
diff --git a/docs/metrics/prometheus/list.md b/docs/metrics/prometheus/list.md
index da9bf793b..4fafae5fe 100644
--- a/docs/metrics/prometheus/list.md
+++ b/docs/metrics/prometheus/list.md
@@ -37,6 +37,7 @@ These metrics can be from any MinIO server once per collection.
 |`minio_node_io_read_bytes`                      |Total bytes read by the process from the underlying storage system, /proc/[pid]/io read_bytes                                |
 |`minio_node_io_wchar_bytes`                     |Total bytes written by the process to the underlying storage system including page cache, /proc/[pid]/io wchar               |
 |`minio_node_io_write_bytes`                     |Total bytes written by the process to the underlying storage system, /proc/[pid]/io write_bytes                              |
+|`minio_node_process_starttime_seconds`          |Start time for MinIO process per node in seconds.                                                                            |
 |`minio_node_syscall_read_total`                 |Total read SysCalls to the kernel. /proc/[pid]/io syscr                                                                      |
 |`minio_node_syscall_write_total`                |Total write SysCalls to the kernel. /proc/[pid]/io syscw                                                                     |
 |`minio_s3_requests_error_total`                 |Total number S3 requests with errors                                                                                         |