Cleanup cgroups collector (#2414)

* Correctly name collector file.
* Fix cgroup summary type as gauge.
* Use a boolean metric rather than a label for enabled.

Signed-off-by: Ben Kochie <superq@gmail.com>

README.md

@@ -193,6 +193,7 @@ from debugfs. And example usage of this would be
 Name     | Description | OS
 ---------|-------------|----
 buddyinfo | Exposes statistics of memory fragments as reported by /proc/buddyinfo. | Linux
+cgroups | A summary of the number of active and enabled cgroups | Linux
 devstat | Exposes device statistics | Dragonfly, FreeBSD
 drbd | Exposes Distributed Replicated Block Device statistics (to version 8.4) | Linux
 ethtool | Exposes network interface information and network driver statistics equivalent to `ethtool`, `ethtool -S`, and `ethtool -i`. | Linux

collector/cgroups_linux.go

@@ -18,21 +18,23 @@ package collector
 
 import (
 	"fmt"
-	"strconv"
 
 	"github.com/go-kit/log"
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/procfs"
 )
 
+const cgroupsCollectorSubsystem = "cgroups"
+
 type cgroupSummaryCollector struct {
 	fs      procfs.FS
 	cgroups *prometheus.Desc
+	enabled *prometheus.Desc
 	logger  log.Logger
 }
 
 func init() {
-	registerCollector("cgroupSummary", defaultEnabled, NewCgroupSummaryCollector)
+	registerCollector(cgroupsCollectorSubsystem, defaultDisabled, NewCgroupSummaryCollector)
 }
 
 // NewCgroupSummaryCollector returns a new Collector exposing a summary of cgroups.
@@ -44,9 +46,14 @@ func NewCgroupSummaryCollector(logger log.Logger) (Collector, error) {
 	return &cgroupSummaryCollector{
 		fs: fs,
 		cgroups: prometheus.NewDesc(
-			prometheus.BuildFQName(namespace, "", "cgroups_total"),
+			prometheus.BuildFQName(namespace, cgroupsCollectorSubsystem, "cgroups"),
 			"Current cgroup number of the subsystem.",
-			[]string{"subsys_name", "enabled"}, nil,
+			[]string{"subsys_name"}, nil,
+		),
+		enabled: prometheus.NewDesc(
+			prometheus.BuildFQName(namespace, cgroupsCollectorSubsystem, "enabled"),
+			"Current cgroup number of the subsystem.",
+			[]string{"subsys_name"}, nil,
 		),
 		logger: logger,
 	}, nil
@@ -59,7 +66,8 @@ func (c *cgroupSummaryCollector) Update(ch chan<- prometheus.Metric) error {
 		return err
 	}
 	for _, cs := range cgroupSummarys {
-		ch <- prometheus.MustNewConstMetric(c.cgroups, prometheus.CounterValue, float64(cs.Cgroups), cs.SubsysName, strconv.Itoa(cs.Enabled))
+		ch <- prometheus.MustNewConstMetric(c.cgroups, prometheus.GaugeValue, float64(cs.Cgroups), cs.SubsysName)
+		ch <- prometheus.MustNewConstMetric(c.enabled, prometheus.GaugeValue, float64(cs.Enabled), cs.SubsysName)
 	}
 	return nil
 }

collector/fixtures/e2e-output.txt

@@ -296,20 +296,34 @@ node_buddyinfo_blocks{node="0",size="8",zone="Normal"} 0
 node_buddyinfo_blocks{node="0",size="9",zone="DMA"} 1
 node_buddyinfo_blocks{node="0",size="9",zone="DMA32"} 0
 node_buddyinfo_blocks{node="0",size="9",zone="Normal"} 0
-# HELP node_cgroups_total Current cgroup number of the subsystem.
+# HELP node_cgroups_cgroups Current cgroup number of the subsystem.
-# TYPE node_cgroups_total counter
+# TYPE node_cgroups_cgroups gauge
-node_cgroups_total{enabled="1",subsys_name="blkio"} 170
+node_cgroups_cgroups{subsys_name="blkio"} 170
-node_cgroups_total{enabled="1",subsys_name="cpu"} 172
+node_cgroups_cgroups{subsys_name="cpu"} 172
-node_cgroups_total{enabled="1",subsys_name="cpuacct"} 172
+node_cgroups_cgroups{subsys_name="cpuacct"} 172
-node_cgroups_total{enabled="1",subsys_name="cpuset"} 47
+node_cgroups_cgroups{subsys_name="cpuset"} 47
-node_cgroups_total{enabled="1",subsys_name="devices"} 170
+node_cgroups_cgroups{subsys_name="devices"} 170
-node_cgroups_total{enabled="1",subsys_name="freezer"} 47
+node_cgroups_cgroups{subsys_name="freezer"} 47
-node_cgroups_total{enabled="1",subsys_name="hugetlb"} 47
+node_cgroups_cgroups{subsys_name="hugetlb"} 47
-node_cgroups_total{enabled="1",subsys_name="memory"} 234
+node_cgroups_cgroups{subsys_name="memory"} 234
-node_cgroups_total{enabled="1",subsys_name="net_cls"} 47
+node_cgroups_cgroups{subsys_name="net_cls"} 47
-node_cgroups_total{enabled="1",subsys_name="perf_event"} 47
+node_cgroups_cgroups{subsys_name="perf_event"} 47
-node_cgroups_total{enabled="1",subsys_name="pids"} 170
+node_cgroups_cgroups{subsys_name="pids"} 170
-node_cgroups_total{enabled="1",subsys_name="rdma"} 1
+node_cgroups_cgroups{subsys_name="rdma"} 1
+# HELP node_cgroups_enabled Current cgroup number of the subsystem.
+# TYPE node_cgroups_enabled gauge
+node_cgroups_enabled{subsys_name="blkio"} 1
+node_cgroups_enabled{subsys_name="cpu"} 1
+node_cgroups_enabled{subsys_name="cpuacct"} 1
+node_cgroups_enabled{subsys_name="cpuset"} 1
+node_cgroups_enabled{subsys_name="devices"} 1
+node_cgroups_enabled{subsys_name="freezer"} 1
+node_cgroups_enabled{subsys_name="hugetlb"} 1
+node_cgroups_enabled{subsys_name="memory"} 1
+node_cgroups_enabled{subsys_name="net_cls"} 1
+node_cgroups_enabled{subsys_name="perf_event"} 1
+node_cgroups_enabled{subsys_name="pids"} 1
+node_cgroups_enabled{subsys_name="rdma"} 1
 # HELP node_context_switches_total Total number of context switches.
 # TYPE node_context_switches_total counter
 node_context_switches_total 3.8014093e+07
@@ -3045,7 +3059,7 @@ node_scrape_collector_success{collector="bcache"} 1
 node_scrape_collector_success{collector="bonding"} 1
 node_scrape_collector_success{collector="btrfs"} 1
 node_scrape_collector_success{collector="buddyinfo"} 1
-node_scrape_collector_success{collector="cgroupSummary"} 1
+node_scrape_collector_success{collector="cgroups"} 1
 node_scrape_collector_success{collector="conntrack"} 1
 node_scrape_collector_success{collector="cpu"} 1
 node_scrape_collector_success{collector="cpufreq"} 1

end-to-end-test.sh

@@ -8,6 +8,7 @@ enabled_collectors=$(cat << COLLECTORS
   btrfs
   buddyinfo
   conntrack
+  cgroups
   cpu
   cpufreq
   diskstats