Adding upstream version 1.34.4.

Signed-off-by: Daniel Baumann <daniel@debian.org>

plugins/inputs/filestat/README.md

@@ -0,0 +1,53 @@
+# File statistics Input Plugin
+
+This plugin gathers metrics about file existence, size, and other file
+statistics.
+
+⭐ Telegraf v0.13.0
+🏷️ system
+💻 all
+
+## Global configuration options <!-- @/docs/includes/plugin_config.md -->
+
+In addition to the plugin-specific configuration settings, plugins support
+additional global and plugin configuration settings. These settings are used to
+modify metrics, tags, and field or create aliases and configure ordering, etc.
+See the [CONFIGURATION.md][CONFIGURATION.md] for more details.
+
+[CONFIGURATION.md]: ../../../docs/CONFIGURATION.md#plugins
+
+## Configuration
+
+```toml @sample.conf
+# Read stats about given file(s)
+[[inputs.filestat]]
+  ## Files to gather stats about.
+  ## These accept standard unix glob matching rules, but with the addition of
+  ## ** as a "super asterisk". See https://github.com/gobwas/glob.
+  files = ["/etc/telegraf/telegraf.conf", "/var/log/**.log"]
+
+  ## If true, read the entire file and calculate an md5 checksum.
+  md5 = false
+```
+
+## Metrics
+
+### Measurements & Fields
+
+- filestat
+  - exists (int, 0 | 1)
+  - size_bytes (int, bytes)
+  - modification_time (int, unix time nanoseconds)
+  - md5 (optional, string)
+
+### Tags
+
+- All measurements have the following tags:
+  - file (the path the to file, as specified in the config)
+
+## Example Output
+
+```text
+filestat,file=/tmp/foo/bar,host=tyrion exists=0i 1507218518192154351
+filestat,file=/Users/sparrc/ws/telegraf.conf,host=tyrion exists=1i,size=47894i,modification_time=1507152973123456789i  1507218518192154351
+```

plugins/inputs/filestat/filestat.go

@@ -0,0 +1,141 @@
+//go:generate ../../../tools/readme_config_includer/generator
+package filestat
+
+import (
+	"crypto/md5" //nolint:gosec // G501: Blocklisted import crypto/md5: weak cryptographic primitive - md5 hash is what is desired in this case
+	_ "embed"
+	"encoding/hex"
+	"io"
+	"os"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/internal/globpath"
+	"github.com/influxdata/telegraf/plugins/inputs"
+)
+
+//go:embed sample.conf
+var sampleConfig string
+
+type FileStat struct {
+	Md5   bool     `toml:"md5"`
+	Files []string `toml:"files"`
+
+	Log telegraf.Logger `toml:"-"`
+
+	// maps full file paths to globmatch obj
+	globs map[string]*globpath.GlobPath
+
+	// files that were missing - we only log the first time it's not found.
+	missingFiles map[string]bool
+	// files that had an error in Stat - we only log the first error.
+	filesWithErrors map[string]bool
+}
+
+func (*FileStat) SampleConfig() string {
+	return sampleConfig
+}
+
+func (f *FileStat) Gather(acc telegraf.Accumulator) error {
+	var err error
+
+	for _, filepath := range f.Files {
+		// Get the compiled glob object for this filepath
+		g, ok := f.globs[filepath]
+		if !ok {
+			if g, err = globpath.Compile(filepath); err != nil {
+				acc.AddError(err)
+				continue
+			}
+			f.globs[filepath] = g
+		}
+
+		files := g.Match()
+		if len(files) == 0 {
+			acc.AddFields("filestat",
+				map[string]interface{}{
+					"exists": int64(0),
+				},
+				map[string]string{
+					"file": filepath,
+				})
+			continue
+		}
+
+		for _, fileName := range files {
+			tags := map[string]string{
+				"file": fileName,
+			}
+			fields := map[string]interface{}{
+				"exists": int64(1),
+			}
+			fileInfo, err := os.Stat(fileName)
+			if os.IsNotExist(err) {
+				fields["exists"] = int64(0)
+				acc.AddFields("filestat", fields, tags)
+				if !f.missingFiles[fileName] {
+					f.Log.Warnf("File %q not found", fileName)
+					f.missingFiles[fileName] = true
+				}
+				continue
+			}
+			f.missingFiles[fileName] = false
+
+			if fileInfo == nil {
+				if !f.filesWithErrors[fileName] {
+					f.filesWithErrors[fileName] = true
+					f.Log.Errorf("Unable to get info for file %q: %v",
+						fileName, err)
+				}
+			} else {
+				f.filesWithErrors[fileName] = false
+				fields["size_bytes"] = fileInfo.Size()
+				fields["modification_time"] = fileInfo.ModTime().UnixNano()
+			}
+
+			if f.Md5 {
+				md5Hash, err := getMd5(fileName)
+				if err != nil {
+					acc.AddError(err)
+				} else {
+					fields["md5_sum"] = md5Hash
+				}
+			}
+
+			acc.AddFields("filestat", fields, tags)
+		}
+	}
+
+	return nil
+}
+
+// Read given file and calculate a md5 hash.
+func getMd5(file string) (string, error) {
+	of, err := os.Open(file)
+	if err != nil {
+		return "", err
+	}
+	defer of.Close()
+
+	//nolint:gosec // G401: Use of weak cryptographic primitive - md5 hash is what is desired in this case
+	hash := md5.New()
+	_, err = io.Copy(hash, of)
+	if err != nil {
+		// fatal error
+		return "", err
+	}
+	return hex.EncodeToString(hash.Sum(nil)), nil
+}
+
+func newFileStat() *FileStat {
+	return &FileStat{
+		globs:           make(map[string]*globpath.GlobPath),
+		missingFiles:    make(map[string]bool),
+		filesWithErrors: make(map[string]bool),
+	}
+}
+
+func init() {
+	inputs.Add("filestat", func() telegraf.Input {
+		return newFileStat()
+	})
+}

plugins/inputs/filestat/filestat_test.go

@@ -0,0 +1,211 @@
+//go:build !windows
+
+// TODO: Windows - should be enabled for Windows when super asterisk is fixed on Windows
+// https://github.com/influxdata/telegraf/issues/6248
+
+package filestat
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+
+	"github.com/influxdata/telegraf/testutil"
+)
+
+var testdataDir = getTestdataDir()
+
+func TestGatherNoMd5(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Files = []string{
+		filepath.Join(testdataDir, "log1.log"),
+		filepath.Join(testdataDir, "log2.log"),
+		filepath.Join(testdataDir, "non_existent_file"),
+	}
+
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	tags1 := map[string]string{
+		"file": filepath.Join(testdataDir, "log1.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags1, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags1, "exists", int64(1)))
+
+	tags2 := map[string]string{
+		"file": filepath.Join(testdataDir, "log2.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags2, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags2, "exists", int64(1)))
+
+	tags3 := map[string]string{
+		"file": filepath.Join(testdataDir, "non_existent_file"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags3, "exists", int64(0)))
+}
+
+func TestGatherExplicitFiles(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Md5 = true
+	fs.Files = []string{
+		filepath.Join(testdataDir, "log1.log"),
+		filepath.Join(testdataDir, "log2.log"),
+		filepath.Join(testdataDir, "non_existent_file"),
+	}
+
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	tags1 := map[string]string{
+		"file": filepath.Join(testdataDir, "log1.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags1, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags1, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags1, "md5_sum", "d41d8cd98f00b204e9800998ecf8427e"))
+
+	tags2 := map[string]string{
+		"file": filepath.Join(testdataDir, "log2.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags2, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags2, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags2, "md5_sum", "d41d8cd98f00b204e9800998ecf8427e"))
+
+	tags3 := map[string]string{
+		"file": filepath.Join(testdataDir, "non_existent_file"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags3, "exists", int64(0)))
+}
+
+func TestNonExistentFile(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Md5 = true
+	fs.Files = []string{
+		"/non/existant/file",
+	}
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	acc.AssertContainsFields(t, "filestat", map[string]interface{}{"exists": int64(0)})
+	require.False(t, acc.HasField("filestat", "error"))
+	require.False(t, acc.HasField("filestat", "md5_sum"))
+	require.False(t, acc.HasField("filestat", "size_bytes"))
+	require.False(t, acc.HasField("filestat", "modification_time"))
+}
+
+func TestGatherGlob(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Md5 = true
+	fs.Files = []string{
+		filepath.Join(testdataDir, "*.log"),
+	}
+
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	tags1 := map[string]string{
+		"file": filepath.Join(testdataDir, "log1.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags1, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags1, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags1, "md5_sum", "d41d8cd98f00b204e9800998ecf8427e"))
+
+	tags2 := map[string]string{
+		"file": filepath.Join(testdataDir, "log2.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags2, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags2, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags2, "md5_sum", "d41d8cd98f00b204e9800998ecf8427e"))
+}
+
+func TestGatherSuperAsterisk(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Md5 = true
+	fs.Files = []string{
+		filepath.Join(testdataDir, "**"),
+	}
+
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	tags1 := map[string]string{
+		"file": filepath.Join(testdataDir, "log1.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags1, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags1, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags1, "md5_sum", "d41d8cd98f00b204e9800998ecf8427e"))
+
+	tags2 := map[string]string{
+		"file": filepath.Join(testdataDir, "log2.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags2, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags2, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags2, "md5_sum", "d41d8cd98f00b204e9800998ecf8427e"))
+
+	tags3 := map[string]string{
+		"file": filepath.Join(testdataDir, "test.conf"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags3, "size_bytes", int64(104)))
+	require.True(t, acc.HasPoint("filestat", tags3, "exists", int64(1)))
+	require.True(t, acc.HasPoint("filestat", tags3, "md5_sum", "5a7e9b77fa25e7bb411dbd17cf403c1f"))
+}
+
+func TestModificationTime(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Files = []string{
+		filepath.Join(testdataDir, "log1.log"),
+	}
+
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	tags1 := map[string]string{
+		"file": filepath.Join(testdataDir, "log1.log"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags1, "size_bytes", int64(0)))
+	require.True(t, acc.HasPoint("filestat", tags1, "exists", int64(1)))
+	require.True(t, acc.HasInt64Field("filestat", "modification_time"))
+}
+
+func TestNoModificationTime(t *testing.T) {
+	fs := newFileStat()
+	fs.Log = testutil.Logger{}
+	fs.Files = []string{
+		filepath.Join(testdataDir, "non_existent_file"),
+	}
+
+	acc := testutil.Accumulator{}
+	require.NoError(t, acc.GatherError(fs.Gather))
+
+	tags1 := map[string]string{
+		"file": filepath.Join(testdataDir, "non_existent_file"),
+	}
+	require.True(t, acc.HasPoint("filestat", tags1, "exists", int64(0)))
+	require.False(t, acc.HasInt64Field("filestat", "modification_time"))
+}
+
+func TestGetMd5(t *testing.T) {
+	md5, err := getMd5(filepath.Join(testdataDir, "test.conf"))
+	require.NoError(t, err)
+	require.Equal(t, "5a7e9b77fa25e7bb411dbd17cf403c1f", md5)
+
+	_, err = getMd5("/tmp/foo/bar/fooooo")
+	require.Error(t, err)
+}
+
+func getTestdataDir() string {
+	dir, err := os.Getwd()
+	if err != nil {
+		// if we cannot even establish the test directory, further progress is meaningless
+		panic(err)
+	}
+
+	return filepath.Join(dir, "testdata")
+}

plugins/inputs/filestat/sample.conf

@@ -0,0 +1,9 @@
+# Read stats about given file(s)
+[[inputs.filestat]]
+  ## Files to gather stats about.
+  ## These accept standard unix glob matching rules, but with the addition of
+  ## ** as a "super asterisk". See https://github.com/gobwas/glob.
+  files = ["/etc/telegraf/telegraf.conf", "/var/log/**.log"]
+
+  ## If true, read the entire file and calculate an md5 checksum.
+  md5 = false

plugins/inputs/filestat/testdata/test.conf

@@ -0,0 +1,5 @@
+# this is a fake testing config file
+# for testing the filestat plugin
+
+option1 = "foo"
+option2 = "bar"