Merge pull request #1445 from davemcphee/coolingDevice

Scrape cooling_device state

CHANGELOG.md

@@ -24,8 +24,9 @@
 * [BUGFIX] Renamed label `state` to `name` on `node_systemd_service_restart_total`. #1393
 * [BUGFIX] Fix netdev nil reference on Darwin #1414
 * [BUGFIX] Strip path.rootfs from mountpoint labels #1421
-* [FEATURE] Add new thermal_zone collector #1425
 * [BUGFIX] Fix empty string in path.rootfs #1464
+* [FEATURE] Add new thermal_zone collector #1425
+* [FEATURE] Add new cooling_device metrics to thermal zone collector #1445
 
 ## 0.18.1 / 2019-06-04
 

README.md

@@ -55,6 +55,7 @@ schedstat | Exposes task scheduler statistics from `/proc/schedstat`. | Linux
 sockstat | Exposes various statistics from `/proc/net/sockstat`. | Linux
 stat | Exposes various statistics from `/proc/stat`. This includes boot time, forks and interrupts. | Linux
 textfile | Exposes statistics read from local disk. The `--collector.textfile.directory` flag must be set. | _any_
+thermal\_zone | Exposes thermal zone & cooling device statistics from `/sys/class/thermal`. | Linux
 time | Exposes the current system time. | _any_
 timex | Exposes selected adjtimex(2) system call stats. | Linux
 uname | Exposes system information as provided by the uname system call. | Darwin, FreeBSD, Linux, OpenBSD

collector/fixtures/e2e-64k-page-output.txt

@@ -178,6 +178,12 @@ node_buddyinfo_blocks{node="0",size="9",zone="Normal"} 0
 # HELP node_context_switches_total Total number of context switches.
 # TYPE node_context_switches_total counter
 node_context_switches_total 3.8014093e+07
+# HELP node_cooling_device_cur_state Current throttle state of the cooling device
+# TYPE node_cooling_device_cur_state gauge
+node_cooling_device_cur_state{name="0",type="Processor"} 0
+# HELP node_cooling_device_max_state Maximum throttle state of the cooling device
+# TYPE node_cooling_device_max_state gauge
+node_cooling_device_max_state{name="0",type="Processor"} 3
 # HELP node_cpu_core_throttles_total Number of times this cpu core has been throttled.
 # TYPE node_cpu_core_throttles_total counter
 node_cpu_core_throttles_total{core="0",package="0"} 5

collector/fixtures/e2e-output.txt

@@ -178,6 +178,12 @@ node_buddyinfo_blocks{node="0",size="9",zone="Normal"} 0
 # HELP node_context_switches_total Total number of context switches.
 # TYPE node_context_switches_total counter
 node_context_switches_total 3.8014093e+07
+# HELP node_cooling_device_cur_state Current throttle state of the cooling device
+# TYPE node_cooling_device_cur_state gauge
+node_cooling_device_cur_state{name="0",type="Processor"} 0
+# HELP node_cooling_device_max_state Maximum throttle state of the cooling device
+# TYPE node_cooling_device_max_state gauge
+node_cooling_device_max_state{name="0",type="Processor"} 3
 # HELP node_cpu_core_throttles_total Number of times this cpu core has been throttled.
 # TYPE node_cpu_core_throttles_total counter
 node_cpu_core_throttles_total{core="0",package="0"} 5

collector/fixtures/sys.ttar

@@ -926,6 +926,9 @@ Mode: 755
 Path: sys/class/thermal/thermal_zone0
 SymlinkTo: ../../devices/virtual/thermal/thermal_zone0
 # ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+Path: sys/class/thermal/cooling_device0
+SymlinkTo: ../../devices/virtual/thermal/cooling_device0
+# ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
 Directory: sys/devices
 Mode: 755
 # ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
@@ -2472,6 +2475,24 @@ Mode: 755
 Directory: sys/devices/virtual/thermal
 Mode: 755
 # ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+Directory: sys/devices/virtual/thermal/cooling_device0
+Mode: 755
+# ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+Path: sys/devices/virtual/thermal/cooling_device0/cur_state
+Lines: 1
+0
+Mode: 644
+# ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+Path: sys/devices/virtual/thermal/cooling_device0/max_state
+Lines: 1
+3
+Mode: 644
+# ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+Path: sys/devices/virtual/thermal/cooling_device0/type
+Lines: 1
+Processor
+Mode: 644
+# ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
 Directory: sys/devices/virtual/thermal/thermal_zone0
 Mode: 755
 # ttar - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

collector/thermal_zone_linux.go

@@ -22,9 +22,14 @@ import (
 	"github.com/prometheus/procfs/sysfs"
 )
 
+const coolingDevice = "cooling_device"
+const thermalZone = "thermal_zone"
+
 type thermalZoneCollector struct {
-	fs       sysfs.FS
-	zoneTemp *prometheus.Desc
+	fs                    sysfs.FS
+	coolingDeviceCurState *prometheus.Desc
+	coolingDeviceMaxState *prometheus.Desc
+	zoneTemp              *prometheus.Desc
 }
 
 func init() {
@@ -41,10 +46,20 @@ func NewThermalZoneCollector() (Collector, error) {
 	return &thermalZoneCollector{
 		fs: fs,
 		zoneTemp: prometheus.NewDesc(
-			prometheus.BuildFQName(namespace, "thermal_zone", "temp"),
+			prometheus.BuildFQName(namespace, thermalZone, "temp"),
 			"Zone temperature in Celsius",
 			[]string{"zone", "type"}, nil,
 		),
+		coolingDeviceCurState: prometheus.NewDesc(
+			prometheus.BuildFQName(namespace, coolingDevice, "cur_state"),
+			"Current throttle state of the cooling device",
+			[]string{"name", "type"}, nil,
+		),
+		coolingDeviceMaxState: prometheus.NewDesc(
+			prometheus.BuildFQName(namespace, coolingDevice, "max_state"),
+			"Maximum throttle state of the cooling device",
+			[]string{"name", "type"}, nil,
+		),
 	}, nil
 }
 
@@ -64,5 +79,28 @@ func (c *thermalZoneCollector) Update(ch chan<- prometheus.Metric) error {
 		)
 	}
 
+	coolingDevices, err := c.fs.ClassCoolingDeviceStats()
+	if err != nil {
+		return err
+	}
+
+	for _, stats := range coolingDevices {
+		ch <- prometheus.MustNewConstMetric(
+			c.coolingDeviceCurState,
+			prometheus.GaugeValue,
+			float64(stats.CurState),
+			stats.Name,
+			stats.Type,
+		)
+
+		ch <- prometheus.MustNewConstMetric(
+			c.coolingDeviceMaxState,
+			prometheus.GaugeValue,
+			float64(stats.MaxState),
+			stats.Name,
+			stats.Type,
+		)
+	}
+
 	return nil
 }