Merge pull request #592 from mallenLF/mssql-perflib

Use perflib to gather metrics in the mssql collector

CODE_OF_CONDUCT.md

@@ -0,0 +1,3 @@
+## Prometheus Community Code of Conduct
+
+Prometheus follows the [CNCF Code of Conduct](https://github.com/cncf/foundation/blob/master/code-of-conduct.md).

MAINTAINERS.md

@@ -1,4 +1,5 @@
 Contributors in alphabetical order
 
+* [Brian Brazil](https://github.com/brian-brazil)
 * [Martin Lindhe](https://github.com/martinlindhe)
 * [Calle Pettersson](https://github.com/carlpett)

README.md

@@ -14,8 +14,10 @@ Name     | Description | Enabled by default
 [cpu](docs/collector.cpu.md) | CPU usage | ✓
 [cs](docs/collector.cs.md) | "Computer System" metrics (system properties, num cpus/total memory) | ✓
 [container](docs/collector.container.md) | Container metrics |
+[dhcp](docs/collector.dhcp.md) | DHCP Server |
 [dns](docs/collector.dns.md) | DNS Server |
 [exchange](docs/collector.exchange.md) | Exchange metrics |
+[fsrmquota](docs/collector.fsrmquota.md) | Microsoft File Server Resource Manager (FSRM) Quotas collector |
 [hyperv](docs/collector.hyperv.md) | Hyper-V hosts |
 [iis](docs/collector.iis.md) | IIS sites and applications |
 [logical_disk](docs/collector.logical_disk.md) | Logical disks, disk I/O | ✓
@@ -45,6 +47,19 @@ Name     | Description | Enabled by default
 
 See the linked documentation on each collector for more information on reported metrics, configuration settings and usage examples.
 
+## Flags
+
+windows_exporter accepts flags to configure certain behaviours. The ones configuring the global behaviour of the exporter are listed below, while collector-specific ones are documented in the respective collector documentation above.
+
+Flag     | Description | Default value
+---------|-------------|--------------------
+`--telemetry.addr` | host:port for exporter. | `:9182`
+`--telemetry.path` | URL path for surfacing collected metrics. | `/metrics`
+`--telemetry.max-requests` | Maximum number of concurrent requests. 0 to disable. | `5`
+`--collectors.enabled` | Comma-separated list of collectors to use. Use `[defaults]` as a placeholder for all the collectors enabled by default." | `[defaults]`
+`--collectors.print` | If true, print available collectors and exit. | 
+`--scrape.timeout-margin` | Seconds to subtract from the timeout allowed by the client. Tune to allow for overhead or high loads. | `0.5`
+
 ## Installation
 The latest release can be downloaded from the [releases page](https://github.com/prometheus-community/windows_exporter/releases).
 
@@ -87,7 +102,7 @@ See [open issues](https://github.com/prometheus-community/windows_exporter/issue
     go get -u github.com/prometheus/promu
     go get -u github.com/prometheus-community/windows_exporter
     cd $env:GOPATH/src/github.com/prometheus-community/windows_exporter
-    promu build -v .
+    promu build -v
     .\windows_exporter.exe
 
 The prometheus metrics will be exposed on [localhost:9182](http://localhost:9182)

collector/collector.go

@@ -58,6 +58,10 @@ var (
 
 func registerCollector(name string, builder collectorBuilder, perfCounterNames ...string) {
 	builders[name] = builder
+	addPerfCounterDependencies(name, perfCounterNames)
+}
+
+func addPerfCounterDependencies(name string, perfCounterNames []string) {
 	perfIndicies := make([]string, 0, len(perfCounterNames))
 	for _, cn := range perfCounterNames {
 		perfIndicies = append(perfIndicies, MapCounterToIndex(cn))
@@ -109,3 +113,9 @@ func PrepareScrapeContext(collectors []string) (*ScrapeContext, error) {
 
 	return &ScrapeContext{objs}, nil
 }
+func boolToFloat(b bool) float64 {
+	if b {
+		return 1.0
+	}
+	return 0.0
+}

collector/dhcp.go

@@ -134,7 +134,7 @@ func NewDhcpCollector() (Collector, error) {
 			nil,
 		),
 		DeniedDueToNonMatch: prometheus.NewDesc(
-			prometheus.BuildFQName(Namespace, subsystem, "denied_due_to_match_total"),
+			prometheus.BuildFQName(Namespace, subsystem, "denied_due_to_nonmatch_total"),
 			"Total number of DHCP requests denied, based on non-matches from the Allow list (DeniedDueToNonMatch)",
 			nil,
 			nil,

collector/fsrmquota.go

@@ -0,0 +1,187 @@
+package collector
+
+import (
+	"github.com/StackExchange/wmi"
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/common/log"
+)
+
+func init() {
+	registerCollector("fsrmquota", newFSRMQuotaCollector)
+}
+
+type FSRMQuotaCollector struct {
+	QuotasCount *prometheus.Desc
+	Path        *prometheus.Desc
+	PeakUsage   *prometheus.Desc
+	Size        *prometheus.Desc
+	Usage       *prometheus.Desc
+
+	Description     *prometheus.Desc
+	Disabled        *prometheus.Desc
+	MatchesTemplate *prometheus.Desc
+	SoftLimit       *prometheus.Desc
+	Template        *prometheus.Desc
+}
+
+func newFSRMQuotaCollector() (Collector, error) {
+	const subsystem = "fsrmquota"
+	return &FSRMQuotaCollector{
+		QuotasCount: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "count"),
+			"Number of Quotas",
+			nil,
+			nil,
+		),
+		PeakUsage: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "peak_usage_bytes"),
+			"The highest amount of disk space usage charged to this quota. (PeakUsage)",
+			[]string{"path", "template"},
+			nil,
+		),
+		Size: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "size_bytes"),
+			"The size of the quota. (Size)",
+			[]string{"path", "template"},
+			nil,
+		),
+		Usage: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "usage_bytes"),
+			"The current amount of disk space usage charged to this quota. (Usage)",
+			[]string{"path", "template"},
+			nil,
+		),
+		Description: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "description"),
+			"Description of the quota (Description)",
+			[]string{"path", "template", "description"},
+			nil,
+		),
+		Disabled: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "disabled"),
+			"If 1, the quota is disabled. The default value is 0. (Disabled)",
+			[]string{"path", "template"},
+			nil,
+		),
+		SoftLimit: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "softlimit"),
+			"If 1, the quota is a soft limit. If 0, the quota is a hard limit. The default value is 0. Optional (SoftLimit)",
+			[]string{"path", "template"},
+			nil,
+		),
+		Template: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "template"),
+			"Quota template name. (Template)",
+			[]string{"path", "template"},
+			nil,
+		),
+		MatchesTemplate: prometheus.NewDesc(
+			prometheus.BuildFQName(Namespace, subsystem, "matchestemplate"),
+			"If 1, the property values of this quota match those values of the template from which it was derived. (MatchesTemplate)",
+			[]string{"path", "template"},
+			nil,
+		),
+	}, nil
+}
+
+// Collect sends the metric values for each metric
+// to the provided prometheus Metric channel.
+func (c *FSRMQuotaCollector) Collect(ctx *ScrapeContext, ch chan<- prometheus.Metric) error {
+	if desc, err := c.collect(ch); err != nil {
+		log.Error("failed collecting fsrmquota metrics:", desc, err)
+		return err
+	}
+	return nil
+}
+
+// MSFT_FSRMQuota docs:
+// https://docs.microsoft.com/en-us/previous-versions/windows/desktop/fsrm/msft-fsrmquota
+type MSFT_FSRMQuota struct {
+	Name string
+
+	Path        string
+	PeakUsage   uint64
+	Size        uint64
+	Usage       uint64
+	Description string
+	Template    string
+	//Threshold             string
+	Disabled        bool
+	MatchesTemplate bool
+	SoftLimit       bool
+}
+
+func (c *FSRMQuotaCollector) collect(ch chan<- prometheus.Metric) (*prometheus.Desc, error) {
+	var dst []MSFT_FSRMQuota
+	q := queryAll(&dst)
+
+	var count int
+
+	if err := wmi.QueryNamespace(q, &dst, "root/microsoft/windows/fsrm"); err != nil {
+		return nil, err
+	}
+
+	for _, quota := range dst {
+
+		count++
+		path := quota.Path
+		template := quota.Template
+		Description := quota.Description
+
+		ch <- prometheus.MustNewConstMetric(
+			c.PeakUsage,
+			prometheus.GaugeValue,
+			float64(quota.PeakUsage),
+			path,
+			template,
+		)
+		ch <- prometheus.MustNewConstMetric(
+			c.Size,
+			prometheus.GaugeValue,
+			float64(quota.Size),
+			path,
+			template,
+		)
+		ch <- prometheus.MustNewConstMetric(
+			c.Usage,
+			prometheus.GaugeValue,
+			float64(quota.Usage),
+			path,
+			template,
+		)
+		ch <- prometheus.MustNewConstMetric(
+			c.Description,
+			prometheus.GaugeValue,
+			1.0,
+			path, template, Description,
+		)
+		ch <- prometheus.MustNewConstMetric(
+			c.Disabled,
+			prometheus.GaugeValue,
+			boolToFloat(quota.Disabled),
+			path,
+			template,
+		)
+		ch <- prometheus.MustNewConstMetric(
+			c.MatchesTemplate,
+			prometheus.GaugeValue,
+			boolToFloat(quota.MatchesTemplate),
+			path,
+			template,
+		)
+		ch <- prometheus.MustNewConstMetric(
+			c.SoftLimit,
+			prometheus.GaugeValue,
+			boolToFloat(quota.SoftLimit),
+			path,
+			template,
+		)
+	}
+
+	ch <- prometheus.MustNewConstMetric(
+		c.QuotasCount,
+		prometheus.GaugeValue,
+		float64(count),
+	)
+	return nil, nil
+}

collector/service.go

@@ -45,7 +45,7 @@ func NewserviceCollector() (Collector, error) {
 		Information: prometheus.NewDesc(
 			prometheus.BuildFQName(Namespace, subsystem, "info"),
 			"A metric with a constant '1' value labeled with service information",
-			[]string{"name", "display_name", "process_id"},
+			[]string{"name", "display_name", "process_id", "run_as"},
 			nil,
 		),
 		State: prometheus.NewDesc(
@@ -89,6 +89,7 @@ type Win32_Service struct {
 	State       string
 	Status      string
 	StartMode   string
+	StartName   *string
 }
 
 var (
@@ -131,9 +132,13 @@ func (c *serviceCollector) collect(ch chan<- prometheus.Metric) (*prometheus.Des
 	if err := wmi.Query(q, &dst); err != nil {
 		return nil, err
 	}
-
 	for _, service := range dst {
 		pid := strconv.FormatUint(uint64(service.ProcessId), 10)
+
+		runAs := ""
+		if service.StartName != nil {
+			runAs = *service.StartName
+		}
 		ch <- prometheus.MustNewConstMetric(
 			c.Information,
 			prometheus.GaugeValue,
@@ -141,6 +146,7 @@ func (c *serviceCollector) collect(ch chan<- prometheus.Metric) (*prometheus.Des
 			strings.ToLower(service.Name),
 			service.DisplayName,
 			pid,
+			runAs,
 		)
 
 		for _, state := range allStates {

docs/collector.fsrmquota.md

@@ -0,0 +1,52 @@
+# Microsoft File Server Resource Manager (FSRM) Quotas collector
+
+The fsrmquota collector exposes metrics about File Server Ressource Manager Quotas. Note that this collector has only been tested against Windows server 2012R2.
+Other FSRM versions may work but are not tested.
+
+|||
+-|-
+Metric name prefix  | `fsrmquota`
+Data source         | wmi
+Counters            | `FSRMQUOTA`
+Enabled by default? | No
+
+## Flags
+
+None
+
+## Metrics
+
+Name | Description | Type | Labels
+-----|-------------|------|-------
+
+`windows_fsrmquota_count` | Number of Quotas | counter |None
+`windows_fsrmquota_description` | A string up to 1KB in size. Optional. The default value is an empty string. (Description) | counter |`path`, `template`,`description`
+`windows_fsrmquota_disabled` | If 1, the quota is disabled. The default value is 0. (Disabled) | counter |`path`, `template`
+`windows_fsrmquota_matchestemplate` | If 1, the property values of this quota match those values of the template from which it was derived. (MatchesTemplate) | counter |`path`, `template`
+`windows_fsrmquota_peak_usage_bytes ` | The highest amount of disk space usage charged to this quota. (PeakUsage) | counter |`path`, `template`
+`windows_fsrmquota_size_bytes` | The size of the quota. If the Template property is not provided then the Size property must be provided (Size) | counter |`path`, `template`
+`windows_fsrmquota_softlimit` | If 1, the quota is a soft limit. If 0, the quota is a hard limit. The default value is 0. Optional (SoftLimit) | counter |`path`, `template`
+`windows_fsrmquota_template` | A valid quota template name. Up to 1KB in size. Optional (Template) | counter |`path`, `template`
+`windows_fsrmquota_usage_bytes` | The current amount of disk space usage charged to this quota. (Usage) | counter |`path`, `template`
+
+
+### Example metric
+Show rate of Quotas usage:
+```
+rate(windows_fsrmquota_usage_bytes)[1d]
+```
+
+## Useful queries
+
+## Alerting examples
+**prometheus.rules**
+```yaml
+  - alert: "HighQuotasUsage"
+    expr: "windows_fsrmquota_usage_bytes{instance="SERVER1.COM:9182"} / windows_fsrmquota_size{instance="SERVER1.COM:9182"} >0.85"
+    for: "10m"
+    labels:
+      severity: "high"
+    annotations:
+      summary: "High Quotas Usage"
+      description: "High use of File Ressource.\n Quotas: {{ $labels.path }}\n Current use : {{ $value }}"
+```

docs/collector.hyperv.md

@@ -96,7 +96,18 @@ Name | Description | Type | Labels
 _This collector does not yet have explained examples, we would appreciate your help adding them!_
 
 ## Useful queries
-_This collector does not yet have any useful queries added, we would appreciate your help adding them!_
+Percent of physical CPU resources used per VM (on instance "localhost")
+```
+(sum (rate(windows_hyperv_vm_cpu_hypervisor_run_time{instance="localhost"}[1m]))) / ignoring(vm) group_left max (windows_cs_logical_processors{instance="localhost"}) / 100000
+```
+Percent of physical CPU resources used by all VMs (on all monitored hosts)
+```
+(sum by (instance)(rate(windows_hyperv_vm_cpu_total_run_time{}[1m]))) / max by (instance)(windows_cs_logical_processors{}) / 100000
+```
+Percent of physical CPU resources by the hosts themselves (on all monitored hosts)
+```
+(sum by (instance)(rate(windows_hyperv_host_cpu_total_run_time{}[1m]))) / sum by (instance)(windows_cs_logical_processors{}) / 100000
+```
 
 ## Alerting examples
 _This collector does not yet have alerting examples, we would appreciate your help adding them!_

docs/collector.service.md

@@ -20,12 +20,12 @@ Example: `--collector.service.services-where="Name='windows_exporter'"`
 
 Name | Description | Type | Labels
 -----|-------------|------|-------
-`windows_service_info` | Contains service information in labels, constant 1 | gauge | name, display_name, process_id
+`windows_service_info` | Contains service information in labels, constant 1 | gauge | name, display_name, process_id, run_as
 `windows_service_state` | The state of the service, 1 if the current state, 0 otherwise | gauge | name, state
 `windows_service_start_mode` | The start mode of the service, 1 if the current start mode, 0 otherwise | gauge | name, start_mode
 `windows_service_status` | The status of the service, 1 if the current status, 0 otherwise | gauge | name, status
 
-For the values of the `state`, `start_mode` and `status` labels, see below.
+For the values of the `state`, `start_mode`, `status` and `run_as` labels, see below.
 
 ### States
 
@@ -66,6 +66,13 @@ A service can have any of the following statuses:
 
 Note that there is some overlap with service state.
 
+### Run As
+
+Account name under which a service runs. Depending on the service type, the account name may be in the form of "DomainName\Username" or UPN format ("Username@DomainName").
+
+It corresponds to the `StartName` attribute of the `Win32_Service` class.
+`StartName` attribute can be NULL and in such case the label is reported as an empty string. Notice that if the attribute is NULL the service is logged on as the `LocalSystem` account or, for kernel or system-level drive, it runs with a default object name created by the I/O system based on the service name, for example, DWDOM\Admin.
+
 ### Example metric
 Lists the services that have a 'disabled' start mode.
 ```

exporter.go

@@ -3,6 +3,7 @@
 package main
 
 import (
+	"encoding/json"
 	"fmt"
 	"net/http"
 	_ "net/http/pprof"
@@ -28,6 +29,17 @@ type windowsCollector struct {
 	collectors        map[string]collector.Collector
 }
 
+// Same struct prometheus uses for their /version endpoint.
+// Separate copy to avoid pulling all of prometheus as a dependency
+type prometheusVersion struct {
+	Version   string `json:"version"`
+	Revision  string `json:"revision"`
+	Branch    string `json:"branch"`
+	BuildUser string `json:"buildUser"`
+	BuildDate string `json:"buildDate"`
+	GoVersion string `json:"goVersion"`
+}
+
 const (
 	defaultCollectors            = "cpu,cs,logical_disk,net,os,service,system,textfile"
 	defaultCollectorsPlaceholder = "[defaults]"
@@ -337,6 +349,21 @@ func main() {
 
 	http.HandleFunc(*metricsPath, withConcurrencyLimit(*maxRequests, h.ServeHTTP))
 	http.HandleFunc("/health", healthCheck)
+	http.HandleFunc("/version", func(w http.ResponseWriter, r *http.Request) {
+		// we can't use "version" directly as it is a package, and not an object that
+		// can be serialized.
+		err := json.NewEncoder(w).Encode(prometheusVersion{
+			Version:   version.Version,
+			Revision:  version.Revision,
+			Branch:    version.Branch,
+			BuildUser: version.BuildUser,
+			BuildDate: version.BuildDate,
+			GoVersion: version.GoVersion,
+		})
+		if err != nil {
+			http.Error(w, fmt.Sprintf("error encoding JSON: %s", err), http.StatusInternalServerError)
+		}
+	})
 	http.HandleFunc("/", func(w http.ResponseWriter, r *http.Request) {
 		_, _ = w.Write([]byte(`<html>
 <head><title>windows_exporter</title></head>