chore(inputs.procstat): Convert bool tagging settings to a list (#14333)

migrations/inputs_procstat/migration.go

@@ -57,6 +57,63 @@ func migrate(tbl *ast.Table) ([]byte, string, error) {
 		delete(plugin, "supervisor_unit")
 	}
 
+	// The tagging options both need the 'tag_with' setting
+	var tagwith []string
+	newTagWith, found := plugin["tag_with"]
+	if found {
+		ntw, ok := newTagWith.([]interface{})
+		if !ok {
+			return nil, "", fmt.Errorf("setting 'tag_with' has wrong type %T", newTagWith)
+		}
+		for _, raw := range ntw {
+			s, ok := raw.(string)
+			if !ok {
+				return nil, "", fmt.Errorf("setting 'tag_with' contains wrong type %T", raw)
+			}
+			if !choice.Contains(s, tagwith) {
+				tagwith = append(tagwith, s)
+			}
+		}
+	}
+
+	// Tagging with PID
+	if oldPidTag, found := plugin["pid_tag"]; found {
+		applied = true
+
+		pt, ok := oldPidTag.(bool)
+		if !ok {
+			return nil, "", fmt.Errorf("setting 'pid_tag' has wrong type %T", oldPidTag)
+		}
+
+		// Add the pid-tagging to 'tag_with' if requested
+		if pt && !choice.Contains("pid", tagwith) {
+			tagwith = append(tagwith, "pid")
+			plugin["tag_with"] = tagwith
+		}
+
+		// Remove deprecated setting
+		delete(plugin, "pid_tag")
+	}
+
+	// Tagging with command-line
+	if oldCmdlinedTag, found := plugin["cmdline_tag"]; found {
+		applied = true
+
+		ct, ok := oldCmdlinedTag.(bool)
+		if !ok {
+			return nil, "", fmt.Errorf("setting 'cmdline_tag' has wrong type %T", oldCmdlinedTag)
+		}
+
+		// Add the pid-tagging to 'tag_with' if requested
+		if ct && !choice.Contains("cmdline", tagwith) {
+			tagwith = append(tagwith, "cmdline")
+			plugin["tag_with"] = tagwith
+		}
+
+		// Remove deprecated setting
+		delete(plugin, "cmdline_tag")
+	}
+
 	// No options migrated so we can exit early
 	if !applied {
 		return nil, "", migrations.ErrNotApplicable

migrations/inputs_procstat/migration_test.go

@@ -9,99 +9,12 @@ import (
 
 	"github.com/influxdata/telegraf/config"
 	_ "github.com/influxdata/telegraf/migrations/inputs_procstat" // register migration
-	_ "github.com/influxdata/telegraf/plugins/inputs/procstat"    // register plugin
+	"github.com/influxdata/telegraf/plugins/inputs/procstat"      // register plugin
 )
 
 func TestNoMigration(t *testing.T) {
-	defaultCfg := []byte(`
-  # Read metrics from MQTT topic(s)
-  [[inputs.mqtt_consumer]]
-  ## Broker URLs for the MQTT server or cluster.  To connect to multiple
-  ## clusters or standalone servers, use a separate plugin instance.
-  ##   example: servers = ["tcp://localhost:1883"]
-  ##            servers = ["ssl://localhost:1883"]
-  ##            servers = ["ws://localhost:1883"]
-  servers = ["tcp://127.0.0.1:1883"]
-
-  ## Topics that will be subscribed to.
-  topics = [
-  "telegraf/host01/cpu",
-  "telegraf/+/mem",
-  "sensors/#",
-  ]
-
-  ## The message topic will be stored in a tag specified by this value.  If set
-  ## to the empty string no topic tag will be created.
-  # topic_tag = "topic"
-
-  ## QoS policy for messages
-  ##   0 = at most once
-  ##   1 = at least once
-  ##   2 = exactly once
-  ##
-  ## When using a QoS of 1 or 2, you should enable persistent_session to allow
-  ## resuming unacknowledged messages.
-  # qos = 0
-
-  ## Connection timeout for initial connection in seconds
-  # connection_timeout = "30s"
-
-  ## Max undelivered messages
-  ## This plugin uses tracking metrics, which ensure messages are read to
-  ## outputs before acknowledging them to the original broker to ensure data
-  ## is not lost. This option sets the maximum messages to read from the
-  ## broker that have not been written by an output.
-  ##
-  ## This value needs to be picked with awareness of the agent's
-  ## metric_batch_size value as well. Setting max undelivered messages too high
-  ## can result in a constant stream of data batches to the output. While
-  ## setting it too low may never flush the broker's messages.
-  # max_undelivered_messages = 1000
-
-  ## Persistent session disables clearing of the client session on connection.
-  ## In order for this option to work you must also set client_id to identify
-  ## the client.  To receive messages that arrived while the client is offline,
-  ## also set the qos option to 1 or 2 and don't forget to also set the QoS when
-  ## publishing.
-  # persistent_session = false
-
-  ## If unset, a random client ID will be generated.
-  # client_id = ""
-
-  ## Username and password to connect MQTT server.
-  # username = "telegraf"
-  # password = "metricsmetricsmetricsmetrics"
-
-  ## Optional TLS Config
-  # tls_ca = "/etc/telegraf/ca.pem"
-  # tls_cert = "/etc/telegraf/cert.pem"
-  # tls_key = "/etc/telegraf/key.pem"
-  ## Use TLS but skip chain & host verification
-  # insecure_skip_verify = false
-
-  ## Client trace messages
-  ## When set to true, and debug mode enabled in the agent settings, the MQTT
-  ## client's messages are included in telegraf logs. These messages are very
-  ## noisey, but essential for debugging issues.
-  # client_trace = false
-
-  ## Data format to consume.
-  ## Each data format has its own unique set of configuration options, read
-  ## more about them here:
-  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
-  data_format = "influx"
-
-  ## Enable extracting tag values from MQTT topics
-  ## _ denotes an ignored entry in the topic path
-  # [[inputs.mqtt_consumer.topic_parsing]]
-  #   topic = ""
-  #   measurement = ""
-  #   tags = ""
-  #   fields = ""
-  ## Value supported is int, float, unit
-  #   [[inputs.mqtt_consumer.topic.types]]
-  #      key = type
-`)
+	plugin := procstat.Procstat{}
+	defaultCfg := []byte(plugin.SampleConfig())
 
 	// Migrate and check that nothing changed
 	output, n, err := config.ApplyMigrations(defaultCfg)

migrations/inputs_procstat/testcases/deprecated_tagging_false/expected.conf

@@ -0,0 +1,3 @@
+[[inputs.procstat]]
+pid_file = "/var/run/nginx.pid"
+supervisor_units = ["webserver", "proxy"]

migrations/inputs_procstat/testcases/deprecated_tagging_false/telegraf.conf

@@ -0,0 +1,47 @@
+# Monitor process cpu and memory usage
+[[inputs.procstat]]
+  ## PID file to monitor process
+  pid_file = "/var/run/nginx.pid"
+  ## executable name (ie, pgrep <exe>)
+  # exe = "nginx"
+  ## pattern as argument for pgrep (ie, pgrep -f <pattern>)
+  # pattern = "nginx"
+  ## user as argument for pgrep (ie, pgrep -u <user>)
+  # user = "nginx"
+  ## Systemd unit name, supports globs when include_systemd_children is set to true
+  # systemd_unit = "nginx.service"
+  # include_systemd_children = false
+  ## CGroup name or path, supports globs
+  # cgroup = "systemd/system.slice/nginx.service"
+  ## Supervisor service names of hypervisorctl management
+  supervisor_unit = ["webserver", "proxy"]
+
+  ## Windows service name
+  # win_service = ""
+
+  ## override for process_name
+  ## This is optional; default is sourced from /proc/<pid>/status
+  # process_name = "bar"
+
+  ## Field name prefix
+  # prefix = ""
+
+  ## When true add the full cmdline as a tag.
+  cmdline_tag = false
+
+  ## Mode to use when calculating CPU usage. Can be one of 'solaris' or 'irix'.
+  # mode = "irix"
+
+  ## Add the PID as a tag instead of as a field.  When collecting multiple
+  ## processes with otherwise matching tags this setting should be enabled to
+  ## ensure each process has a unique identity.
+  ##
+  ## Enabling this option may result in a large number of series, especially
+  ## when processes have a short lifetime.
+  pid_tag = false
+
+  ## Method to use when finding process IDs.  Can be one of 'pgrep', or
+  ## 'native'.  The pgrep finder calls the pgrep executable in the PATH while
+  ## the native finder performs the search directly in a manor dependent on the
+  ## platform.  Default is 'pgrep'
+  # pid_finder = "pgrep"

migrations/inputs_procstat/testcases/deprecated_tagging_true/expected.conf

@@ -0,0 +1,4 @@
+[[inputs.procstat]]
+pid_file = "/var/run/nginx.pid"
+supervisor_units = ["webserver", "proxy"]
+tag_with = ["pid", "cmdline"]

migrations/inputs_procstat/testcases/deprecated_tagging_true/telegraf.conf

@@ -0,0 +1,47 @@
+# Monitor process cpu and memory usage
+[[inputs.procstat]]
+  ## PID file to monitor process
+  pid_file = "/var/run/nginx.pid"
+  ## executable name (ie, pgrep <exe>)
+  # exe = "nginx"
+  ## pattern as argument for pgrep (ie, pgrep -f <pattern>)
+  # pattern = "nginx"
+  ## user as argument for pgrep (ie, pgrep -u <user>)
+  # user = "nginx"
+  ## Systemd unit name, supports globs when include_systemd_children is set to true
+  # systemd_unit = "nginx.service"
+  # include_systemd_children = false
+  ## CGroup name or path, supports globs
+  # cgroup = "systemd/system.slice/nginx.service"
+  ## Supervisor service names of hypervisorctl management
+  supervisor_unit = ["webserver", "proxy"]
+
+  ## Windows service name
+  # win_service = ""
+
+  ## override for process_name
+  ## This is optional; default is sourced from /proc/<pid>/status
+  # process_name = "bar"
+
+  ## Field name prefix
+  # prefix = ""
+
+  ## When true add the full cmdline as a tag.
+  cmdline_tag = true
+
+  ## Mode to use when calculating CPU usage. Can be one of 'solaris' or 'irix'.
+  # mode = "irix"
+
+  ## Add the PID as a tag instead of as a field.  When collecting multiple
+  ## processes with otherwise matching tags this setting should be enabled to
+  ## ensure each process has a unique identity.
+  ##
+  ## Enabling this option may result in a large number of series, especially
+  ## when processes have a short lifetime.
+  pid_tag = true
+
+  ## Method to use when finding process IDs.  Can be one of 'pgrep', or
+  ## 'native'.  The pgrep finder calls the pgrep executable in the PATH while
+  ## the native finder performs the search directly in a manor dependent on the
+  ## platform.  Default is 'pgrep'
+  # pid_finder = "pgrep"

migrations/inputs_procstat/testcases/deprecated_tagging_true_merge/expected.conf

@@ -0,0 +1,4 @@
+[[inputs.procstat]]
+pid_file = "/var/run/nginx.pid"
+supervisor_units = ["webserver", "proxy"]
+tag_with = ["ppid", "pid", "cmdline"]

migrations/inputs_procstat/testcases/deprecated_tagging_true_merge/telegraf.conf

@@ -0,0 +1,49 @@
+# Monitor process cpu and memory usage
+[[inputs.procstat]]
+  ## PID file to monitor process
+  pid_file = "/var/run/nginx.pid"
+  ## executable name (ie, pgrep <exe>)
+  # exe = "nginx"
+  ## pattern as argument for pgrep (ie, pgrep -f <pattern>)
+  # pattern = "nginx"
+  ## user as argument for pgrep (ie, pgrep -u <user>)
+  # user = "nginx"
+  ## Systemd unit name, supports globs when include_systemd_children is set to true
+  # systemd_unit = "nginx.service"
+  # include_systemd_children = false
+  ## CGroup name or path, supports globs
+  # cgroup = "systemd/system.slice/nginx.service"
+  ## Supervisor service names of hypervisorctl management
+  supervisor_unit = ["webserver", "proxy"]
+
+  ## Windows service name
+  # win_service = ""
+
+  ## override for process_name
+  ## This is optional; default is sourced from /proc/<pid>/status
+  # process_name = "bar"
+
+  ## Field name prefix
+  # prefix = ""
+
+  ## When true add the full cmdline as a tag.
+  cmdline_tag = true
+
+  ## Mode to use when calculating CPU usage. Can be one of 'solaris' or 'irix'.
+  # mode = "irix"
+
+  ## Add the PID as a tag instead of as a field.  When collecting multiple
+  ## processes with otherwise matching tags this setting should be enabled to
+  ## ensure each process has a unique identity.
+  ##
+  ## Enabling this option may result in a large number of series, especially
+  ## when processes have a short lifetime.
+  pid_tag = true
+
+  ## Method to use when finding process IDs.  Can be one of 'pgrep', or
+  ## 'native'.  The pgrep finder calls the pgrep executable in the PATH while
+  ## the native finder performs the search directly in a manor dependent on the
+  ## platform.  Default is 'pgrep'
+  # pid_finder = "pgrep"
+
+  tag_with = ["ppid"]

migrations/inputs_procstat/testcases/deprecated_tagging_true_merge_overlap/expected.conf

@@ -0,0 +1,4 @@
+[[inputs.procstat]]
+pid_file = "/var/run/nginx.pid"
+supervisor_units = ["webserver", "proxy"]
+tag_with = ["ppid", "cmdline", "pid"]

migrations/inputs_procstat/testcases/deprecated_tagging_true_merge_overlap/telegraf.conf

@@ -0,0 +1,49 @@
+# Monitor process cpu and memory usage
+[[inputs.procstat]]
+  ## PID file to monitor process
+  pid_file = "/var/run/nginx.pid"
+  ## executable name (ie, pgrep <exe>)
+  # exe = "nginx"
+  ## pattern as argument for pgrep (ie, pgrep -f <pattern>)
+  # pattern = "nginx"
+  ## user as argument for pgrep (ie, pgrep -u <user>)
+  # user = "nginx"
+  ## Systemd unit name, supports globs when include_systemd_children is set to true
+  # systemd_unit = "nginx.service"
+  # include_systemd_children = false
+  ## CGroup name or path, supports globs
+  # cgroup = "systemd/system.slice/nginx.service"
+  ## Supervisor service names of hypervisorctl management
+  supervisor_unit = ["webserver", "proxy"]
+
+  ## Windows service name
+  # win_service = ""
+
+  ## override for process_name
+  ## This is optional; default is sourced from /proc/<pid>/status
+  # process_name = "bar"
+
+  ## Field name prefix
+  # prefix = ""
+
+  ## When true add the full cmdline as a tag.
+  cmdline_tag = true
+
+  ## Mode to use when calculating CPU usage. Can be one of 'solaris' or 'irix'.
+  # mode = "irix"
+
+  ## Add the PID as a tag instead of as a field.  When collecting multiple
+  ## processes with otherwise matching tags this setting should be enabled to
+  ## ensure each process has a unique identity.
+  ##
+  ## Enabling this option may result in a large number of series, especially
+  ## when processes have a short lifetime.
+  pid_tag = true
+
+  ## Method to use when finding process IDs.  Can be one of 'pgrep', or
+  ## 'native'.  The pgrep finder calls the pgrep executable in the PATH while
+  ## the native finder performs the search directly in a manor dependent on the
+  ## platform.  Default is 'pgrep'
+  # pid_finder = "pgrep"
+
+  tag_with = ["ppid", "cmdline"]

plugins/inputs/procstat/README.md

@@ -55,19 +55,22 @@ See the [CONFIGURATION.md][CONFIGURATION.md] for more details.
   ## Field name prefix
   # prefix = ""
 
-  ## When true add the full cmdline as a tag.
-  # cmdline_tag = false
-
   ## Mode to use when calculating CPU usage. Can be one of 'solaris' or 'irix'.
   # mode = "irix"
 
-  ## Add the PID as a tag instead of as a field.  When collecting multiple
-  ## processes with otherwise matching tags this setting should be enabled to
-  ## ensure each process has a unique identity.
-  ##
-  ## Enabling this option may result in a large number of series, especially
-  ## when processes have a short lifetime.
-  # pid_tag = false
+  ## Add the given information tag instead of a field
+  ## This allows to create unique metrics/series when collecting processes with
+  ## otherwise identical tags. However, please be careful as this can easily
+  ## result in a large number of series, especially with short-lived processes,
+  ## creating high cardinality at the output.
+  ## Available options are:
+  ##   cmdline -- full commandline
+  ##   pid     -- ID of the process
+  ##   ppid    -- ID of the process' parent
+  ##   status  -- state of the process
+  ##   user    -- username owning the process
+  # tag_with = []
+
 
   ## Method to use when finding process IDs.  Can be one of 'pgrep', or
   ## 'native'.  The pgrep finder calls the pgrep executable in the PATH while

plugins/inputs/procstat/process.go

@@ -3,6 +3,7 @@ package procstat
 import (
 	"fmt"
 	"runtime"
+	"strconv"
 	"time"
 
 	"github.com/influxdata/telegraf"
@@ -15,7 +16,7 @@ type Process interface {
 	Name() (string, error)
 	SetTag(string, string)
 	MemoryMaps(bool) (*[]process.MemoryMapsStat, error)
-	Metric(prefix string, cmdLineTag, solarisMode bool) telegraf.Metric
+	Metric(prefix string, tagging map[string]bool, solarisMode bool) telegraf.Metric
 }
 
 type PIDFinder interface {
@@ -64,42 +65,12 @@ func (p *Proc) percent(_ time.Duration) (float64, error) {
 }
 
 // Add metrics a single Process
-func (p *Proc) Metric(prefix string, cmdLineTag, solarisMode bool) telegraf.Metric {
+func (p *Proc) Metric(prefix string, tagging map[string]bool, solarisMode bool) telegraf.Metric {
 	if prefix != "" {
 		prefix += "_"
 	}
 
 	fields := make(map[string]interface{})
-
-	if _, nameInTags := p.tags["process_name"]; !nameInTags {
-		name, err := p.Name()
-		if err == nil {
-			p.tags["process_name"] = name
-		}
-	}
-
-	if _, ok := p.tags["user"]; !ok {
-		user, err := p.Username()
-		if err == nil {
-			p.tags["user"] = user
-		}
-	}
-
-	// If pid is not present as a tag, include it as a field.
-	if _, pidInTags := p.tags["pid"]; !pidInTags {
-		fields["pid"] = p.Pid
-	}
-
-	// Add the command line as a tag if the option is set
-	if cmdLineTag {
-		if _, ok := p.tags["cmdline"]; !ok {
-			cmdline, err := p.Cmdline()
-			if err == nil {
-				p.tags["cmdline"] = cmdline
-			}
-		}
-	}
-
 	numThreads, err := p.NumThreads()
 	if err == nil {
 		fields[prefix+"num_threads"] = numThreads
@@ -206,14 +177,54 @@ func (p *Proc) Metric(prefix string, cmdLineTag, solarisMode bool) telegraf.Metr
 		}
 	}
 
+	// Add the tags as requested by the user
+	cmdline, err := p.Cmdline()
+	if err == nil {
+		if tagging["cmdline"] {
+			p.tags["cmdline"] = cmdline
+		} else {
+			fields[prefix+"cmdline"] = cmdline
+		}
+	}
+
+	if tagging["pid"] {
+		p.tags["pid"] = strconv.Itoa(int(p.Pid))
+	} else {
+		fields["pid"] = p.Pid
+	}
+
 	ppid, err := p.Ppid()
 	if err == nil {
-		fields[prefix+"ppid"] = ppid
+		if tagging["ppid"] {
+			p.tags["ppid"] = strconv.Itoa(int(ppid))
+		} else {
+			fields[prefix+"ppid"] = ppid
+		}
 	}
 
 	status, err := p.Status()
 	if err == nil {
-		fields[prefix+"status"] = status[0]
+		if tagging["status"] {
+			p.tags["status"] = status[0]
+		} else {
+			fields[prefix+"status"] = status[0]
+		}
+	}
+
+	user, err := p.Username()
+	if err == nil {
+		if tagging["user"] {
+			p.tags["user"] = user
+		} else {
+			fields[prefix+"user"] = user
+		}
+	}
+
+	if _, exists := p.tags["process_name"]; !exists {
+		name, err := p.Name()
+		if err == nil {
+			p.tags["process_name"] = name
+		}
 	}
 
 	return metric.New("procstat", p.tags, fields, time.Time{})

plugins/inputs/procstat/procstat.go

@@ -24,6 +24,7 @@ var sampleConfig string
 
 // execCommand is so tests can mock out exec.Command usage.
 var execCommand = exec.Command
+var availableTags = []string{"cmdline", "pid", "ppid", "status", "user"}
 
 type PID int32
 
@@ -33,7 +34,7 @@ type Procstat struct {
 	Exe                    string          `toml:"exe"`
 	Pattern                string          `toml:"pattern"`
 	Prefix                 string          `toml:"prefix"`
-	CmdLineTag             bool            `toml:"cmdline_tag"`
+	CmdLineTag             bool            `toml:"cmdline_tag" deprecated:"1.29.0;use 'tag_with' instead"`
 	ProcessName            string          `toml:"process_name"`
 	User                   string          `toml:"user"`
 	SystemdUnits           string          `toml:"systemd_units"`
@@ -41,14 +42,16 @@ type Procstat struct {
 	SupervisorUnits        []string        `toml:"supervisor_units"`
 	IncludeSystemdChildren bool            `toml:"include_systemd_children"`
 	CGroup                 string          `toml:"cgroup"`
-	PidTag                 bool            `toml:"pid_tag"`
+	PidTag                 bool            `toml:"pid_tag" deprecated:"1.29.0;use 'tag_with' instead"`
 	WinService             string          `toml:"win_service"`
 	Mode                   string          `toml:"mode"`
+	TagWith                []string        `toml:"tag_with"`
 	Log                    telegraf.Logger `toml:"-"`
 
 	solarisMode bool
 	finder      PIDFinder
 	processes   map[PID]Process
+	tagging     map[string]bool
 
 	createProcess func(PID) (Process, error)
 }
@@ -66,6 +69,23 @@ func (p *Procstat) Init() error {
 	// Check solaris mode
 	p.solarisMode = strings.ToLower(p.Mode) == "solaris"
 
+	// Keep the old settings for compatibility
+	if p.PidTag && !choice.Contains("pid", p.TagWith) {
+		p.TagWith = append(p.TagWith, "pid")
+	}
+	if p.CmdLineTag && !choice.Contains("cmdline", p.TagWith) {
+		p.TagWith = append(p.TagWith, "cmdline")
+	}
+
+	// Check tagging and setup LUT
+	if err := choice.CheckSlice(p.TagWith, availableTags); err != nil {
+		return fmt.Errorf("invalid tag_with setting: %w", err)
+	}
+	p.tagging = make(map[string]bool, len(p.TagWith))
+	for _, tag := range p.TagWith {
+		p.tagging[tag] = true
+	}
+
 	// Keep the old settings for compatibility
 	for _, u := range p.SupervisorUnit {
 		if !choice.Contains(u, p.SupervisorUnits) {
@@ -97,7 +117,7 @@ func (p *Procstat) Init() error {
 		// see https://github.com/shirou/gopsutil/blob/v3.23.10/process/process_darwin.go#L235
 		requiresChildren := len(p.SupervisorUnits) > 0 && p.Pattern != ""
 		if requiresChildren && runtime.GOOS == "darwin" {
-			return errors.New("configuration requires the 'pgrep' finder on you OS")
+			return errors.New("configuration requires 'pgrep' finder on your OS")
 		}
 		p.finder = &NativeFinder{}
 	case "test":
@@ -159,17 +179,13 @@ func (p *Procstat) Gather(acc telegraf.Accumulator) error {
 					proc.SetTag(k, v)
 				}
 
-				// Add pid tag if needed
-				if p.PidTag {
-					proc.SetTag("pid", strconv.Itoa(int(pid)))
-				}
 				if p.ProcessName != "" {
 					proc.SetTag("process_name", p.ProcessName)
 				}
 				p.processes[pid] = proc
 			}
 			running[pid] = true
-			m := proc.Metric(p.Prefix, p.CmdLineTag, p.solarisMode)
+			m := proc.Metric(p.Prefix, p.tagging, p.solarisMode)
 			m.SetTime(now)
 			acc.AddMetric(m)
 		}

plugins/inputs/procstat/procstat_test.go

@@ -6,6 +6,7 @@ import (
 	"os/exec"
 	"path/filepath"
 	"runtime"
+	"strconv"
 	"strings"
 	"testing"
 	"time"
@@ -116,8 +117,9 @@ type testProc struct {
 	tags map[string]string
 }
 
-func newTestProc(_ PID) (Process, error) {
+func newTestProc(pid PID) (Process, error) {
 	proc := &testProc{
+		pid:  pid,
 		tags: make(map[string]string),
 	}
 	return proc, nil
@@ -139,14 +141,12 @@ func (p *testProc) MemoryMaps(bool) (*[]process.MemoryMapsStat, error) {
 	return &[]process.MemoryMapsStat{}, nil
 }
 
-func (p *testProc) Metric(prefix string, cmdLineTag, _ bool) telegraf.Metric {
+func (p *testProc) Metric(prefix string, tagging map[string]bool, _ bool) telegraf.Metric {
 	if prefix != "" {
 		prefix += "_"
 	}
 
 	fields := map[string]interface{}{
-		"pid":                                   int32(p.pid),
-		"ppid":                                  int32(0),
 		prefix + "num_fds":                      int32(0),
 		prefix + "num_threads":                  int32(0),
 		prefix + "voluntary_context_switches":   int64(0),
@@ -167,20 +167,46 @@ func (p *testProc) Metric(prefix string, cmdLineTag, _ bool) telegraf.Metric {
 		prefix + "memory_rss":                   uint64(0),
 		prefix + "memory_vms":                   uint64(0),
 		prefix + "memory_usage":                 float32(0),
-		prefix + "status":                       "running",
 	}
 
 	tags := map[string]string{
 		"process_name": "test_proc",
-		"user":         "testuser",
 	}
 	for k, v := range p.tags {
 		tags[k] = v
 	}
 
-	if cmdLineTag {
+	// Add the tags as requested by the user
+	if tagging["cmdline"] {
 		tags["cmdline"] = "test_proc"
+	} else {
+		fields[prefix+"cmdline"] = "test_proc"
 	}
+
+	if tagging["pid"] {
+		tags["pid"] = strconv.Itoa(int(p.pid))
+	} else {
+		fields["pid"] = int32(p.pid)
+	}
+
+	if tagging["ppid"] {
+		tags["ppid"] = "0"
+	} else {
+		fields[prefix+"ppid"] = int32(0)
+	}
+
+	if tagging["status"] {
+		tags["status"] = "running"
+	} else {
+		fields[prefix+"status"] = "running"
+	}
+
+	if tagging["user"] {
+		tags["user"] = "testuser"
+	} else {
+		fields[prefix+"user"] = "testuser"
+	}
+
 	return metric.New("procstat", tags, fields, time.Time{})
 }
 
@@ -207,7 +233,7 @@ func TestInitRequiresChildDarwin(t *testing.T) {
 		PidFinder:       "native",
 		Log:             testutil.Logger{},
 	}
-	require.ErrorContains(t, p.Init(), "requires the 'pgrep' finder")
+	require.ErrorContains(t, p.Init(), "requires 'pgrep' finder")
 }
 
 func TestInitMissingPidMethod(t *testing.T) {

plugins/inputs/procstat/sample.conf

@@ -26,19 +26,22 @@
   ## Field name prefix
   # prefix = ""
 
-  ## When true add the full cmdline as a tag.
-  # cmdline_tag = false
-
   ## Mode to use when calculating CPU usage. Can be one of 'solaris' or 'irix'.
   # mode = "irix"
 
-  ## Add the PID as a tag instead of as a field.  When collecting multiple
-  ## processes with otherwise matching tags this setting should be enabled to
-  ## ensure each process has a unique identity.
-  ##
-  ## Enabling this option may result in a large number of series, especially
-  ## when processes have a short lifetime.
-  # pid_tag = false
+  ## Add the given information tag instead of a field
+  ## This allows to create unique metrics/series when collecting processes with
+  ## otherwise identical tags. However, please be careful as this can easily
+  ## result in a large number of series, especially with short-lived processes,
+  ## creating high cardinality at the output.
+  ## Available options are:
+  ##   cmdline -- full commandline
+  ##   pid     -- ID of the process
+  ##   ppid    -- ID of the process' parent
+  ##   status  -- state of the process
+  ##   user    -- username owning the process
+  # tag_with = []
+
 
   ## Method to use when finding process IDs.  Can be one of 'pgrep', or
   ## 'native'.  The pgrep finder calls the pgrep executable in the PATH while