windows_exporter/collector/process.go

// returns data points from Win32_PerfRawData_PerfProc_Process
// https://msdn.microsoft.com/en-us/library/aa394323(v=vs.85).aspx - Win32_PerfRawData_PerfProc_Process class
package collector

import (
	"bytes"
	"flag"
	"strconv"
	"strings"

	"github.com/StackExchange/wmi"
	"github.com/prometheus/client_golang/prometheus"
	"github.com/prometheus/common/log"
)

func init() {
	Factories["process"] = NewProcessCollector
}

var (
	processWhereClause = flag.String("collector.process.processes-where", "", "WQL 'where' clause to use in WMI metrics query. Limits the response to the processes you specify and reduces the size of the response.")
)

// A ProcessCollector is a Prometheus collector for WMI Win32_PerfRawData_PerfProc_Process metrics
type ProcessCollector struct {
	StartTime         *prometheus.Desc
	CPUTimeTotal      *prometheus.Desc
	HandleCount       *prometheus.Desc
	IOBytesTotal      *prometheus.Desc
	IOOperationsTotal *prometheus.Desc
	PageFaultsTotal   *prometheus.Desc
	PageFileBytes     *prometheus.Desc
	PoolBytes         *prometheus.Desc
	PriorityBase      *prometheus.Desc
	PrivateBytes      *prometheus.Desc
	ThreadCount       *prometheus.Desc
	VirtualBytes      *prometheus.Desc
	WorkingSet        *prometheus.Desc

	queryWhereClause string
}

// NewProcessCollector ...
func NewProcessCollector() (Collector, error) {
	const subsystem = "process"

	var wc bytes.Buffer
	if *processWhereClause != "" {
		wc.WriteString("WHERE ")
		wc.WriteString(*processWhereClause)
	} else {
		log.Warn("No where-clause specified for process collector. This will generate a very large number of metrics!")
	}

	return &ProcessCollector{
		StartTime: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "start_time"),
			"Time of process start.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		CPUTimeTotal: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "cpu_time_total"),
			"Returns elapsed time that all of the threads of this process used the processor to execute instructions by mode (privileged, user). An instruction is the basic unit of execution in a computer, a thread is the object that executes instructions, and a process is the object created when a program is run. Code executed to handle some hardware interrupts and trap conditions is included in this count.",
			[]string{"process", "process_id", "creating_process_id", "mode"},
			nil,
		),
		HandleCount: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "handle_count"),
			"Total number of handles the process has open. This number is the sum of the handles currently open by each thread in the process.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		IOBytesTotal: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "io_bytes_total"),
			"Bytes issued to I/O operations in different modes (read, write, other). This property counts all I/O activity generated by the process to include file, network, and device I/Os. Read and write mode includes data operations; other mode includes those that do not involve data, such as control operations. ",
			[]string{"process", "process_id", "creating_process_id", "mode"},
			nil,
		),
		IOOperationsTotal: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "io_operations_total"),
			"I/O operations issued in different modes (read, write, other). This property counts all I/O activity generated by the process to include file, network, and device I/Os. Read and write mode includes data operations; other mode includes those that do not involve data, such as control operations. ",
			[]string{"process", "process_id", "creating_process_id", "mode"},
			nil,
		),
		PageFaultsTotal: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "page_faults_total"),
			"Page faults by the threads executing in this process. A page fault occurs when a thread refers to a virtual memory page that is not in its working set in main memory. This can cause the page not to be fetched from disk if it is on the standby list and hence already in main memory, or if it is in use by another process with which the page is shared.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		PageFileBytes: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "page_file_bytes"),
			"Current number of bytes this process has used in the paging file(s). Paging files are used to store pages of memory used by the process that are not contained in other files. Paging files are shared by all processes, and lack of space in paging files can prevent other processes from allocating memory.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		PoolBytes: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "pool_bytes"),
			"Pool Bytes is the last observed number of bytes in the paged or nonpaged pool. The nonpaged pool is an area of system memory (physical memory used by the operating system) for objects that cannot be written to disk, but must remain in physical memory as long as they are allocated. The paged pool is an area of system memory (physical memory used by the operating system) for objects that can be written to disk when they are not being used. Nonpaged pool bytes is calculated differently than paged pool bytes, so it might not equal the total of paged pool bytes.",
			[]string{"process", "process_id", "creating_process_id", "pool"},
			nil,
		),
		PriorityBase: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "priority_base"),
			"Current base priority of this process. Threads within a process can raise and lower their own base priority relative to the process base priority of the process.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		PrivateBytes: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "private_bytes"),
			"Current number of bytes this process has allocated that cannot be shared with other processes.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		ThreadCount: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "thread_count"),
			"Number of threads currently active in this process. An instruction is the basic unit of execution in a processor, and a thread is the object that executes instructions. Every running process has at least one thread.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		VirtualBytes: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "virtual_bytes"),
			"Current size, in bytes, of the virtual address space that the process is using. Use of virtual address space does not necessarily imply corresponding use of either disk or main memory pages. Virtual space is finite and, by using too much, the process can limit its ability to load libraries.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		WorkingSet: prometheus.NewDesc(
			prometheus.BuildFQName(Namespace, subsystem, "working_set"),
			"Maximum number of bytes in the working set of this process at any point in time. The working set is the set of memory pages touched recently by the threads in the process. If free memory in the computer is above a threshold, pages are left in the working set of a process even if they are not in use. When free memory falls below a threshold, pages are trimmed from working sets. If they are needed, they are then soft-faulted back into the working set before they leave main memory.",
			[]string{"process", "process_id", "creating_process_id"},
			nil,
		),
		queryWhereClause: wc.String(),
	}, nil
}

// Collect sends the metric values for each metric
// to the provided prometheus Metric channel.
func (c *ProcessCollector) Collect(ch chan<- prometheus.Metric) error {
	if desc, err := c.collect(ch); err != nil {
		log.Errorln("[ERROR] failed collecting process metrics:", desc, err)
		return err
	}
	return nil
}

type Win32_PerfRawData_PerfProc_Process struct {
	Name                    string
	CreatingProcessID       uint32
	ElapsedTime             uint64
	Frequency_Object        uint64
	HandleCount             uint32
	IDProcess               uint32
	IODataBytesPersec       uint64
	IODataOperationsPersec  uint64
	IOOtherBytesPersec      uint64
	IOOtherOperationsPersec uint64
	IOReadBytesPersec       uint64
	IOReadOperationsPersec  uint64
	IOWriteBytesPersec      uint64
	IOWriteOperationsPersec uint64
	PageFaultsPersec        uint32
	PageFileBytes           uint64
	PageFileBytesPeak       uint64
	PercentPrivilegedTime   uint64
	PercentProcessorTime    uint64
	PercentUserTime         uint64
	PoolNonpagedBytes       uint32
	PoolPagedBytes          uint32
	PriorityBase            uint32
	PrivateBytes            uint64
	ThreadCount             uint32
	Timestamp_Object        uint64
	VirtualBytes            uint64
	VirtualBytesPeak        uint64
	WorkingSet              uint64
	WorkingSetPeak          uint64
	WorkingSetPrivate       uint64
}

func (c *ProcessCollector) collect(ch chan<- prometheus.Metric) (*prometheus.Desc, error) {
	var dst []Win32_PerfRawData_PerfProc_Process
	q := wmi.CreateQuery(&dst, c.queryWhereClause)
	if err := wmi.Query(q, &dst); err != nil {
		return nil, err
	}

	for _, process := range dst {

		if process.Name == "_Total" {
			continue
		}
		// Duplicate processes are suffixed # and an index number. Remove those.
		processName := strings.Split(process.Name, "#")[0]
		pid := strconv.FormatUint(uint64(process.IDProcess), 10)
		cpid := strconv.FormatUint(uint64(process.CreatingProcessID), 10)

		ch <- prometheus.MustNewConstMetric(
			c.StartTime,
			prometheus.GaugeValue,
			// convert from Windows timestamp (1 jan 1601) to unix timestamp (1 jan 1970)
			float64(process.ElapsedTime-116444736000000000)/float64(process.Frequency_Object),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.HandleCount,
			prometheus.GaugeValue,
			float64(process.HandleCount),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.CPUTimeTotal,
			prometheus.CounterValue,
			float64(process.PercentPrivilegedTime)*ticksToSecondsScaleFactor,
			processName,
			pid,
			cpid,
			"privileged",
		)

		ch <- prometheus.MustNewConstMetric(
			c.CPUTimeTotal,
			prometheus.CounterValue,
			float64(process.PercentUserTime)*ticksToSecondsScaleFactor,
			processName,
			pid,
			cpid,
			"user",
		)

		ch <- prometheus.MustNewConstMetric(
			c.IOBytesTotal,
			prometheus.CounterValue,
			float64(process.IOOtherBytesPersec),
			processName,
			pid,
			cpid,
			"other",
		)

		ch <- prometheus.MustNewConstMetric(
			c.IOOperationsTotal,
			prometheus.CounterValue,
			float64(process.IOOtherOperationsPersec),
			processName,
			pid,
			cpid,
			"other",
		)

		ch <- prometheus.MustNewConstMetric(
			c.IOBytesTotal,
			prometheus.CounterValue,
			float64(process.IOReadBytesPersec),
			processName,
			pid,
			cpid,
			"read",
		)

		ch <- prometheus.MustNewConstMetric(
			c.IOOperationsTotal,
			prometheus.CounterValue,
			float64(process.IOReadOperationsPersec),
			processName,
			pid,
			cpid,
			"read",
		)

		ch <- prometheus.MustNewConstMetric(
			c.IOBytesTotal,
			prometheus.CounterValue,
			float64(process.IOWriteBytesPersec),
			processName,
			pid,
			cpid,
			"write",
		)

		ch <- prometheus.MustNewConstMetric(
			c.IOOperationsTotal,
			prometheus.CounterValue,
			float64(process.IOWriteOperationsPersec),
			processName,
			pid,
			cpid,
			"write",
		)

		ch <- prometheus.MustNewConstMetric(
			c.PageFaultsTotal,
			prometheus.CounterValue,
			float64(process.PageFaultsPersec),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.PageFileBytes,
			prometheus.GaugeValue,
			float64(process.PageFileBytes),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.PoolBytes,
			prometheus.GaugeValue,
			float64(process.PoolNonpagedBytes),
			processName,
			pid,
			cpid,
			"nonpaged",
		)

		ch <- prometheus.MustNewConstMetric(
			c.PoolBytes,
			prometheus.GaugeValue,
			float64(process.PoolPagedBytes),
			processName,
			pid,
			cpid,
			"paged",
		)

		ch <- prometheus.MustNewConstMetric(
			c.PriorityBase,
			prometheus.GaugeValue,
			float64(process.PriorityBase),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.PrivateBytes,
			prometheus.GaugeValue,
			float64(process.PrivateBytes),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.ThreadCount,
			prometheus.GaugeValue,
			float64(process.ThreadCount),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.VirtualBytes,
			prometheus.GaugeValue,
			float64(process.VirtualBytes),
			processName,
			pid,
			cpid,
		)

		ch <- prometheus.MustNewConstMetric(
			c.WorkingSet,
			prometheus.GaugeValue,
			float64(process.WorkingSet),
			processName,
			pid,
			cpid,
		)
	}

	return nil, nil
}