logparser input plugin

closes #102
closes #328
This commit is contained in:
Cameron Sparr 2016-06-02 18:47:15 +01:00
parent d50a1e83ac
commit cb3c54a1ae
21 changed files with 1713 additions and 123 deletions

View File

@ -5,6 +5,7 @@
- [#1340](https://github.com/influxdata/telegraf/issues/1340): statsd: do not log every dropped metric.
- [#1368](https://github.com/influxdata/telegraf/pull/1368): Add precision rounding to all metrics on collection.
- [#1390](https://github.com/influxdata/telegraf/pull/1390): Add support for Tengine
- [#1320](https://github.com/influxdata/telegraf/pull/1320): Logparser input plugin for parsing grok-style log patterns.
### Bugfixes

1
Godeps
View File

@ -47,6 +47,7 @@ github.com/shirou/gopsutil 586bb697f3ec9f8ec08ffefe18f521a64534037c
github.com/soniah/gosnmp b1b4f885b12c5dcbd021c5cee1c904110de6db7d
github.com/streadway/amqp b4f3ceab0337f013208d31348b578d83c0064744
github.com/stretchr/testify 1f4a1643a57e798696635ea4c126e9127adb7d3c
github.com/vjeantet/grok 83bfdfdfd1a8146795b28e547a8e3c8b28a466c2
github.com/wvanbergen/kafka 46f9a1cf3f670edec492029fadded9c2d9e18866
github.com/wvanbergen/kazoo-go 0f768712ae6f76454f987c3356177e138df258f8
github.com/zensqlmonitor/go-mssqldb ffe5510c6fa5e15e6d983210ab501c815b56b363

View File

@ -1516,6 +1516,35 @@
# data_format = "influx"
# # Stream and parse log file(s).
# [[inputs.logparser]]
# ## Log files to parse.
# ## These accept standard unix glob matching rules, but with the addition of
# ## ** as a "super asterisk". ie:
# ## /var/log/**.log -> recursively find all .log files in /var/log
# ## /var/log/*/*.log -> find all .log files with a parent dir in /var/log
# ## /var/log/apache.log -> only tail the apache log file
# files = ["/var/log/influxdb/influxdb.log"]
# ## Read file from beginning.
# from_beginning = false
#
# ## Parse logstash-style "grok" patterns:
# ## Telegraf built-in parsing patterns: https://goo.gl/dkay10
# [inputs.logparser.grok]
# ## This is a list of patterns to check the given log file(s) for.
# ## Note that adding patterns here increases processing time. The most
# ## efficient configuration is to have one pattern per logparser.
# ## Other common built-in patterns are:
# ## %{COMMON_LOG_FORMAT} (plain apache & nginx access logs)
# ## %{COMBINED_LOG_FORMAT} (access logs + referrer & agent)
# patterns = ["%{INFLUXDB_HTTPD_LOG}"]
# ## Full path(s) to custom pattern files.
# custom_pattern_files = []
# ## Custom patterns can also be defined here. Put one pattern per line.
# custom_patterns = '''
# '''
# # Read metrics from MQTT topic(s)
# [[inputs.mqtt_consumer]]
# servers = ["localhost:1883"]

79
filter/filter.go Normal file
View File

@ -0,0 +1,79 @@
package filter
import (
"strings"
"github.com/gobwas/glob"
)
type Filter interface {
Match(string) bool
}
// CompileFilter takes a list of string filters and returns a Filter interface
// for matching a given string against the filter list. The filter list
// supports glob matching too, ie:
//
// f, _ := CompileFilter([]string{"cpu", "mem", "net*"})
// f.Match("cpu") // true
// f.Match("network") // true
// f.Match("memory") // false
//
func CompileFilter(filters []string) (Filter, error) {
// return if there is nothing to compile
if len(filters) == 0 {
return nil, nil
}
// check if we can compile a non-glob filter
noGlob := true
for _, filter := range filters {
if hasMeta(filter) {
noGlob = false
break
}
}
switch {
case noGlob:
// return non-globbing filter if not needed.
return compileFilterNoGlob(filters), nil
case len(filters) == 1:
return glob.Compile(filters[0])
default:
return glob.Compile("{" + strings.Join(filters, ",") + "}")
}
}
// hasMeta reports whether path contains any magic glob characters.
func hasMeta(s string) bool {
return strings.IndexAny(s, "*?[") >= 0
}
type filter struct {
m map[string]struct{}
}
func (f *filter) Match(s string) bool {
_, ok := f.m[s]
return ok
}
type filtersingle struct {
s string
}
func (f *filtersingle) Match(s string) bool {
return f.s == s
}
func compileFilterNoGlob(filters []string) Filter {
if len(filters) == 1 {
return &filtersingle{s: filters[0]}
}
out := filter{m: make(map[string]struct{})}
for _, filter := range filters {
out.m[filter] = struct{}{}
}
return &out
}

96
filter/filter_test.go Normal file
View File

@ -0,0 +1,96 @@
package filter
import (
"testing"
"github.com/stretchr/testify/assert"
)
func TestCompileFilter(t *testing.T) {
f, err := CompileFilter([]string{})
assert.NoError(t, err)
assert.Nil(t, f)
f, err = CompileFilter([]string{"cpu"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.False(t, f.Match("cpu0"))
assert.False(t, f.Match("mem"))
f, err = CompileFilter([]string{"cpu*"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.True(t, f.Match("cpu0"))
assert.False(t, f.Match("mem"))
f, err = CompileFilter([]string{"cpu", "mem"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.False(t, f.Match("cpu0"))
assert.True(t, f.Match("mem"))
f, err = CompileFilter([]string{"cpu", "mem", "net*"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.False(t, f.Match("cpu0"))
assert.True(t, f.Match("mem"))
assert.True(t, f.Match("network"))
}
var benchbool bool
func BenchmarkFilterSingleNoGlobFalse(b *testing.B) {
f, _ := CompileFilter([]string{"cpu"})
var tmp bool
for n := 0; n < b.N; n++ {
tmp = f.Match("network")
}
benchbool = tmp
}
func BenchmarkFilterSingleNoGlobTrue(b *testing.B) {
f, _ := CompileFilter([]string{"cpu"})
var tmp bool
for n := 0; n < b.N; n++ {
tmp = f.Match("cpu")
}
benchbool = tmp
}
func BenchmarkFilter(b *testing.B) {
f, _ := CompileFilter([]string{"cpu", "mem", "net*"})
var tmp bool
for n := 0; n < b.N; n++ {
tmp = f.Match("network")
}
benchbool = tmp
}
func BenchmarkFilterNoGlob(b *testing.B) {
f, _ := CompileFilter([]string{"cpu", "mem", "net"})
var tmp bool
for n := 0; n < b.N; n++ {
tmp = f.Match("net")
}
benchbool = tmp
}
func BenchmarkFilter2(b *testing.B) {
f, _ := CompileFilter([]string{"aa", "bb", "c", "ad", "ar", "at", "aq",
"aw", "az", "axxx", "ab", "cpu", "mem", "net*"})
var tmp bool
for n := 0; n < b.N; n++ {
tmp = f.Match("network")
}
benchbool = tmp
}
func BenchmarkFilter2NoGlob(b *testing.B) {
f, _ := CompileFilter([]string{"aa", "bb", "c", "ad", "ar", "at", "aq",
"aw", "az", "axxx", "ab", "cpu", "mem", "net"})
var tmp bool
for n := 0; n < b.N; n++ {
tmp = f.Match("net")
}
benchbool = tmp
}

View File

@ -17,8 +17,6 @@ import (
"strings"
"time"
"unicode"
"github.com/gobwas/glob"
)
const alphanum string = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"
@ -209,27 +207,6 @@ func WaitTimeout(c *exec.Cmd, timeout time.Duration) error {
}
}
// CompileFilter takes a list of glob "filters", ie:
// ["MAIN.*", "CPU.*", "NET"]
// and compiles them into a glob object. This glob object can
// then be used to match keys to the filter.
func CompileFilter(filters []string) (glob.Glob, error) {
var out glob.Glob
// return if there is nothing to compile
if len(filters) == 0 {
return out, nil
}
var err error
if len(filters) == 1 {
out, err = glob.Compile(filters[0])
} else {
out, err = glob.Compile("{" + strings.Join(filters, ",") + "}")
}
return out, err
}
// RandomSleep will sleep for a random amount of time up to max.
// If the shutdown channel is closed, it will return before it has finished
// sleeping.

View File

@ -107,37 +107,6 @@ func TestRunError(t *testing.T) {
assert.Error(t, err)
}
func TestCompileFilter(t *testing.T) {
f, err := CompileFilter([]string{})
assert.NoError(t, err)
assert.Nil(t, f)
f, err = CompileFilter([]string{"cpu"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.False(t, f.Match("cpu0"))
assert.False(t, f.Match("mem"))
f, err = CompileFilter([]string{"cpu*"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.True(t, f.Match("cpu0"))
assert.False(t, f.Match("mem"))
f, err = CompileFilter([]string{"cpu", "mem"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.False(t, f.Match("cpu0"))
assert.True(t, f.Match("mem"))
f, err = CompileFilter([]string{"cpu", "mem", "net*"})
assert.NoError(t, err)
assert.True(t, f.Match("cpu"))
assert.False(t, f.Match("cpu0"))
assert.True(t, f.Match("mem"))
assert.True(t, f.Match("network"))
}
func TestRandomSleep(t *testing.T) {
// test that zero max returns immediately
s := time.Now()

View File

@ -3,80 +3,78 @@ package internal_models
import (
"fmt"
"github.com/gobwas/glob"
"github.com/influxdata/telegraf"
"github.com/influxdata/telegraf/internal"
"github.com/influxdata/telegraf/filter"
)
// TagFilter is the name of a tag, and the values on which to filter
type TagFilter struct {
Name string
Filter []string
filter glob.Glob
filter filter.Filter
}
// Filter containing drop/pass and tagdrop/tagpass rules
type Filter struct {
NameDrop []string
nameDrop glob.Glob
nameDrop filter.Filter
NamePass []string
namePass glob.Glob
namePass filter.Filter
FieldDrop []string
fieldDrop glob.Glob
fieldDrop filter.Filter
FieldPass []string
fieldPass glob.Glob
fieldPass filter.Filter
TagDrop []TagFilter
TagPass []TagFilter
TagExclude []string
tagExclude glob.Glob
tagExclude filter.Filter
TagInclude []string
tagInclude glob.Glob
tagInclude filter.Filter
IsActive bool
}
// Compile all Filter lists into glob.Glob objects.
// Compile all Filter lists into filter.Filter objects.
func (f *Filter) CompileFilter() error {
var err error
f.nameDrop, err = internal.CompileFilter(f.NameDrop)
f.nameDrop, err = filter.CompileFilter(f.NameDrop)
if err != nil {
return fmt.Errorf("Error compiling 'namedrop', %s", err)
}
f.namePass, err = internal.CompileFilter(f.NamePass)
f.namePass, err = filter.CompileFilter(f.NamePass)
if err != nil {
return fmt.Errorf("Error compiling 'namepass', %s", err)
}
f.fieldDrop, err = internal.CompileFilter(f.FieldDrop)
f.fieldDrop, err = filter.CompileFilter(f.FieldDrop)
if err != nil {
return fmt.Errorf("Error compiling 'fielddrop', %s", err)
}
f.fieldPass, err = internal.CompileFilter(f.FieldPass)
f.fieldPass, err = filter.CompileFilter(f.FieldPass)
if err != nil {
return fmt.Errorf("Error compiling 'fieldpass', %s", err)
}
f.tagExclude, err = internal.CompileFilter(f.TagExclude)
f.tagExclude, err = filter.CompileFilter(f.TagExclude)
if err != nil {
return fmt.Errorf("Error compiling 'tagexclude', %s", err)
}
f.tagInclude, err = internal.CompileFilter(f.TagInclude)
f.tagInclude, err = filter.CompileFilter(f.TagInclude)
if err != nil {
return fmt.Errorf("Error compiling 'taginclude', %s", err)
}
for i, _ := range f.TagDrop {
f.TagDrop[i].filter, err = internal.CompileFilter(f.TagDrop[i].Filter)
f.TagDrop[i].filter, err = filter.CompileFilter(f.TagDrop[i].Filter)
if err != nil {
return fmt.Errorf("Error compiling 'tagdrop', %s", err)
}
}
for i, _ := range f.TagPass {
f.TagPass[i].filter, err = internal.CompileFilter(f.TagPass[i].Filter)
f.TagPass[i].filter, err = filter.CompileFilter(f.TagPass[i].Filter)
if err != nil {
return fmt.Errorf("Error compiling 'tagpass', %s", err)
}

View File

@ -253,51 +253,6 @@ func TestFilter_TagDrop(t *testing.T) {
}
}
func TestFilter_CompileFilterError(t *testing.T) {
f := Filter{
NameDrop: []string{"", ""},
}
assert.Error(t, f.CompileFilter())
f = Filter{
NamePass: []string{"", ""},
}
assert.Error(t, f.CompileFilter())
f = Filter{
FieldDrop: []string{"", ""},
}
assert.Error(t, f.CompileFilter())
f = Filter{
FieldPass: []string{"", ""},
}
assert.Error(t, f.CompileFilter())
f = Filter{
TagExclude: []string{"", ""},
}
assert.Error(t, f.CompileFilter())
f = Filter{
TagInclude: []string{"", ""},
}
assert.Error(t, f.CompileFilter())
filters := []TagFilter{
TagFilter{
Name: "cpu",
Filter: []string{"{foobar}"},
}}
f = Filter{
TagDrop: filters,
}
require.Error(t, f.CompileFilter())
filters = []TagFilter{
TagFilter{
Name: "cpu",
Filter: []string{"{foobar}"},
}}
f = Filter{
TagPass: filters,
}
require.Error(t, f.CompileFilter())
}
func TestFilter_ShouldMetricsPass(t *testing.T) {
m := testutil.TestMetric(1, "testmetric")
f := Filter{

View File

@ -29,6 +29,7 @@ import (
_ "github.com/influxdata/telegraf/plugins/inputs/jolokia"
_ "github.com/influxdata/telegraf/plugins/inputs/kafka_consumer"
_ "github.com/influxdata/telegraf/plugins/inputs/leofs"
_ "github.com/influxdata/telegraf/plugins/inputs/logparser"
_ "github.com/influxdata/telegraf/plugins/inputs/lustre2"
_ "github.com/influxdata/telegraf/plugins/inputs/mailchimp"
_ "github.com/influxdata/telegraf/plugins/inputs/memcached"

View File

@ -0,0 +1,89 @@
# logparser Input Plugin
The logparser plugin streams and parses the given logfiles. Currently it only
has the capability of parsing "grok" patterns from logfiles, which also supports
regex patterns.
### Configuration:
```toml
[[inputs.logparser]]
## Log files to parse.
## These accept standard unix glob matching rules, but with the addition of
## ** as a "super asterisk". ie:
## /var/log/**.log -> recursively find all .log files in /var/log
## /var/log/*/*.log -> find all .log files with a parent dir in /var/log
## /var/log/apache.log -> only tail the apache log file
files = ["/var/log/influxdb/influxdb.log"]
## Read file from beginning.
from_beginning = false
## Parse logstash-style "grok" patterns:
## Telegraf builtin parsing patterns: https://goo.gl/dkay10
[inputs.logparser.grok]
## This is a list of patterns to check the given log file(s) for.
## Note that adding patterns here increases processing time. The most
## efficient configuration is to have one file & pattern per logparser.
patterns = ["%{INFLUXDB_HTTPD_LOG}"]
## Full path(s) to custom pattern files.
custom_pattern_files = []
## Custom patterns can also be defined here. Put one pattern per line.
custom_patterns = '''
'''
```
## Grok Parser
The grok parser uses a slightly modified version of logstash "grok" patterns,
with the format `%{<capture_syntax>[:<semantic_name>][:<modifier>]}`
Telegraf has many of it's own
[built-in patterns](https://github.com/influxdata/telegraf/blob/master/plugins/inputs/logparser/grok/patterns/influx-patterns),
as well as supporting
[logstash's builtin patterns](https://github.com/logstash-plugins/logstash-patterns-core/blob/master/patterns/grok-patterns).
The best way to get acquainted with grok patterns is to read the logstash docs,
which are available here:
https://www.elastic.co/guide/en/logstash/current/plugins-filters-grok.html
If you need help building patterns to match your logs,
you will find the http://grokdebug.herokuapp.com application quite useful!
By default all named captures are converted into string fields.
Modifiers can be used to convert captures to other types or tags.
Timestamp modifiers can be used to convert captures to the timestamp of the
parsed metric.
- Available modifiers:
- string (default if nothing is specified)
- int
- float
- duration (ie, 5.23ms gets converted to int nanoseconds)
- tag (converts the field into a tag)
- drop (drops the field completely)
- Timestamp modifiers:
- ts-ansic ("Mon Jan _2 15:04:05 2006")
- ts-unix ("Mon Jan _2 15:04:05 MST 2006")
- ts-ruby ("Mon Jan 02 15:04:05 -0700 2006")
- ts-rfc822 ("02 Jan 06 15:04 MST")
- ts-rfc822z ("02 Jan 06 15:04 -0700")
- ts-rfc850 ("Monday, 02-Jan-06 15:04:05 MST")
- ts-rfc1123 ("Mon, 02 Jan 2006 15:04:05 MST")
- ts-rfc1123z ("Mon, 02 Jan 2006 15:04:05 -0700")
- ts-rfc3339 ("2006-01-02T15:04:05Z07:00")
- ts-rfc3339nano ("2006-01-02T15:04:05.999999999Z07:00")
- ts-httpd ("02/Jan/2006:15:04:05 -0700")
- ts-epoch (seconds since unix epoch)
- ts-epochnano (nanoseconds since unix epoch)
- ts-"CUSTOM"
CUSTOM time layouts must be within quotes and be the representation of the
"reference time", which is `Mon Jan 2 15:04:05 -0700 MST 2006`
See https://golang.org/pkg/time/#Parse for more details.

View File

@ -0,0 +1,373 @@
package grok
import (
"bufio"
"fmt"
"log"
"os"
"regexp"
"strconv"
"strings"
"time"
"github.com/vjeantet/grok"
"github.com/influxdata/telegraf"
)
var timeFormats = map[string]string{
"ts-ansic": "Mon Jan _2 15:04:05 2006",
"ts-unix": "Mon Jan _2 15:04:05 MST 2006",
"ts-ruby": "Mon Jan 02 15:04:05 -0700 2006",
"ts-rfc822": "02 Jan 06 15:04 MST",
"ts-rfc822z": "02 Jan 06 15:04 -0700", // RFC822 with numeric zone
"ts-rfc850": "Monday, 02-Jan-06 15:04:05 MST",
"ts-rfc1123": "Mon, 02 Jan 2006 15:04:05 MST",
"ts-rfc1123z": "Mon, 02 Jan 2006 15:04:05 -0700", // RFC1123 with numeric zone
"ts-rfc3339": "2006-01-02T15:04:05Z07:00",
"ts-rfc3339nano": "2006-01-02T15:04:05.999999999Z07:00",
"ts-httpd": "02/Jan/2006:15:04:05 -0700",
"ts-epoch": "EPOCH",
"ts-epochnano": "EPOCH_NANO",
}
const (
INT = "int"
TAG = "tag"
FLOAT = "float"
STRING = "string"
DURATION = "duration"
DROP = "drop"
)
var (
// matches named captures that contain a type.
// ie,
// %{NUMBER:bytes:int}
// %{IPORHOST:clientip:tag}
// %{HTTPDATE:ts1:ts-http}
// %{HTTPDATE:ts2:ts-"02 Jan 06 15:04"}
typedRe = regexp.MustCompile(`%{\w+:(\w+):(ts-".+"|t?s?-?\w+)}`)
// matches a plain pattern name. ie, %{NUMBER}
patternOnlyRe = regexp.MustCompile(`%{(\w+)}`)
)
type Parser struct {
Patterns []string
CustomPatterns string
CustomPatternFiles []string
// typeMap is a map of patterns -> capture name -> modifier,
// ie, {
// "%{TESTLOG}":
// {
// "bytes": "int",
// "clientip": "tag"
// }
// }
typeMap map[string]map[string]string
// tsMap is a map of patterns -> capture name -> timestamp layout.
// ie, {
// "%{TESTLOG}":
// {
// "httptime": "02/Jan/2006:15:04:05 -0700"
// }
// }
tsMap map[string]map[string]string
// patterns is a map of all of the parsed patterns from CustomPatterns
// and CustomPatternFiles.
// ie, {
// "DURATION": "%{NUMBER}[nuµm]?s"
// "RESPONSE_CODE": "%{NUMBER:rc:tag}"
// }
patterns map[string]string
g *grok.Grok
tsModder *tsModder
}
func (p *Parser) Compile() error {
p.typeMap = make(map[string]map[string]string)
p.tsMap = make(map[string]map[string]string)
p.patterns = make(map[string]string)
p.tsModder = &tsModder{}
var err error
p.g, err = grok.NewWithConfig(&grok.Config{NamedCapturesOnly: true})
if err != nil {
return err
}
p.CustomPatterns = DEFAULT_PATTERNS + p.CustomPatterns
if len(p.CustomPatterns) != 0 {
scanner := bufio.NewScanner(strings.NewReader(p.CustomPatterns))
p.addCustomPatterns(scanner)
}
for _, filename := range p.CustomPatternFiles {
file, err := os.Open(filename)
if err != nil {
return err
}
scanner := bufio.NewScanner(bufio.NewReader(file))
p.addCustomPatterns(scanner)
}
return p.compileCustomPatterns()
}
func (p *Parser) ParseLine(line string) (telegraf.Metric, error) {
var err error
var values map[string]string
// the matching pattern string
var patternName string
for _, pattern := range p.Patterns {
if values, err = p.g.Parse(pattern, line); err != nil {
return nil, err
}
if len(values) != 0 {
patternName = pattern
break
}
}
if len(values) == 0 {
return nil, nil
}
fields := make(map[string]interface{})
tags := make(map[string]string)
timestamp := time.Now()
for k, v := range values {
if k == "" || v == "" {
continue
}
var t string
// check if pattern has some modifiers
if types, ok := p.typeMap[patternName]; ok {
t = types[k]
}
// if we didn't find a modifier, check if we have a timestamp layout
if t == "" {
if ts, ok := p.tsMap[patternName]; ok {
// check if the modifier is a timestamp layout
if layout, ok := ts[k]; ok {
t = layout
}
}
}
// if we didn't find a type OR timestamp modifier, assume string
if t == "" {
t = STRING
}
switch t {
case INT:
iv, err := strconv.ParseInt(v, 10, 64)
if err != nil {
log.Printf("ERROR parsing %s to int: %s", v, err)
} else {
fields[k] = iv
}
case FLOAT:
fv, err := strconv.ParseFloat(v, 64)
if err != nil {
log.Printf("ERROR parsing %s to float: %s", v, err)
} else {
fields[k] = fv
}
case DURATION:
d, err := time.ParseDuration(v)
if err != nil {
log.Printf("ERROR parsing %s to duration: %s", v, err)
} else {
fields[k] = int64(d)
}
case TAG:
tags[k] = v
case STRING:
fields[k] = strings.Trim(v, `"`)
case "EPOCH":
iv, err := strconv.ParseInt(v, 10, 64)
if err != nil {
log.Printf("ERROR parsing %s to int: %s", v, err)
} else {
timestamp = time.Unix(iv, 0)
}
case "EPOCH_NANO":
iv, err := strconv.ParseInt(v, 10, 64)
if err != nil {
log.Printf("ERROR parsing %s to int: %s", v, err)
} else {
timestamp = time.Unix(0, iv)
}
case DROP:
// goodbye!
default:
ts, err := time.Parse(t, v)
if err == nil {
timestamp = ts
} else {
log.Printf("ERROR parsing %s to time layout [%s]: %s", v, t, err)
}
}
}
return telegraf.NewMetric("logparser_grok", tags, fields, p.tsModder.tsMod(timestamp))
}
func (p *Parser) addCustomPatterns(scanner *bufio.Scanner) {
for scanner.Scan() {
line := strings.TrimSpace(scanner.Text())
if len(line) > 0 && line[0] != '#' {
names := strings.SplitN(line, " ", 2)
p.patterns[names[0]] = names[1]
}
}
}
func (p *Parser) compileCustomPatterns() error {
var err error
// check if the pattern contains a subpattern that is already defined
// replace it with the subpattern for modifier inheritance.
for i := 0; i < 2; i++ {
for name, pattern := range p.patterns {
subNames := patternOnlyRe.FindAllStringSubmatch(pattern, -1)
for _, subName := range subNames {
if subPattern, ok := p.patterns[subName[1]]; ok {
pattern = strings.Replace(pattern, subName[0], subPattern, 1)
}
}
p.patterns[name] = pattern
}
}
// check if pattern contains modifiers. Parse them out if it does.
for name, pattern := range p.patterns {
if typedRe.MatchString(pattern) {
// this pattern has modifiers, so parse out the modifiers
pattern, err = p.parseTypedCaptures(name, pattern)
if err != nil {
return err
}
p.patterns[name] = pattern
}
}
return p.g.AddPatternsFromMap(p.patterns)
}
// parseTypedCaptures parses the capture types, and then deletes the type from
// the line so that it is a valid "grok" pattern again.
// ie,
// %{NUMBER:bytes:int} => %{NUMBER:bytes} (stores %{NUMBER}->bytes->int)
// %{IPORHOST:clientip:tag} => %{IPORHOST:clientip} (stores %{IPORHOST}->clientip->tag)
func (p *Parser) parseTypedCaptures(name, pattern string) (string, error) {
matches := typedRe.FindAllStringSubmatch(pattern, -1)
// grab the name of the capture pattern
patternName := "%{" + name + "}"
// create type map for this pattern
p.typeMap[patternName] = make(map[string]string)
p.tsMap[patternName] = make(map[string]string)
// boolean to verify that each pattern only has a single ts- data type.
hasTimestamp := false
for _, match := range matches {
// regex capture 1 is the name of the capture
// regex capture 2 is the type of the capture
if strings.HasPrefix(match[2], "ts-") {
if hasTimestamp {
return pattern, fmt.Errorf("logparser pattern compile error: "+
"Each pattern is allowed only one named "+
"timestamp data type. pattern: %s", pattern)
}
if f, ok := timeFormats[match[2]]; ok {
p.tsMap[patternName][match[1]] = f
} else {
p.tsMap[patternName][match[1]] = strings.TrimSuffix(strings.TrimPrefix(match[2], `ts-"`), `"`)
}
hasTimestamp = true
} else {
p.typeMap[patternName][match[1]] = match[2]
}
// the modifier is not a valid part of a "grok" pattern, so remove it
// from the pattern.
pattern = strings.Replace(pattern, ":"+match[2]+"}", "}", 1)
}
return pattern, nil
}
// tsModder is a struct for incrementing identical timestamps of log lines
// so that we don't push identical metrics that will get overwritten.
type tsModder struct {
dupe time.Time
last time.Time
incr time.Duration
incrn time.Duration
rollover time.Duration
}
// tsMod increments the given timestamp one unit more from the previous
// duplicate timestamp.
// the increment unit is determined as the next smallest time unit below the
// most significant time unit of ts.
// ie, if the input is at ms precision, it will increment it 1µs.
func (t *tsModder) tsMod(ts time.Time) time.Time {
defer func() { t.last = ts }()
// don't mod the time if we don't need to
if t.last.IsZero() || ts.IsZero() {
t.incrn = 0
t.rollover = 0
return ts
}
if !ts.Equal(t.last) && !ts.Equal(t.dupe) {
t.incr = 0
t.incrn = 0
t.rollover = 0
return ts
}
if ts.Equal(t.last) {
t.dupe = ts
}
if ts.Equal(t.dupe) && t.incr == time.Duration(0) {
tsNano := ts.UnixNano()
d := int64(10)
counter := 1
for {
a := tsNano % d
if a > 0 {
break
}
d = d * 10
counter++
}
switch {
case counter <= 6:
t.incr = time.Nanosecond
case counter <= 9:
t.incr = time.Microsecond
case counter > 9:
t.incr = time.Millisecond
}
}
t.incrn++
if t.incrn == 999 && t.incr > time.Nanosecond {
t.rollover = t.incr * t.incrn
t.incrn = 1
t.incr = t.incr / 1000
if t.incr < time.Nanosecond {
t.incr = time.Nanosecond
}
}
return ts.Add(t.incr*t.incrn + t.rollover)
}

View File

@ -0,0 +1,508 @@
package grok
import (
"testing"
"time"
"github.com/influxdata/telegraf"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
var benchM telegraf.Metric
func Benchmark_ParseLine_CommonLogFormat(b *testing.B) {
p := &Parser{
Patterns: []string{"%{COMMON_LOG_FORMAT}"},
}
p.Compile()
var m telegraf.Metric
for n := 0; n < b.N; n++ {
m, _ = p.ParseLine(`127.0.0.1 user-identifier frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326`)
}
benchM = m
}
func Benchmark_ParseLine_CombinedLogFormat(b *testing.B) {
p := &Parser{
Patterns: []string{"%{COMBINED_LOG_FORMAT}"},
}
p.Compile()
var m telegraf.Metric
for n := 0; n < b.N; n++ {
m, _ = p.ParseLine(`127.0.0.1 user-identifier frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326 "-" "Mozilla"`)
}
benchM = m
}
func Benchmark_ParseLine_InfluxLog(b *testing.B) {
p := &Parser{
Patterns: []string{"%{INFLUXDB_HTTPD_LOG}"},
}
p.Compile()
var m telegraf.Metric
for n := 0; n < b.N; n++ {
m, _ = p.ParseLine(`[httpd] 192.168.1.1 - - [14/Jun/2016:11:33:29 +0100] "POST /write?consistency=any&db=telegraf&precision=ns&rp= HTTP/1.1" 204 0 "-" "InfluxDBClient" 6f61bc44-321b-11e6-8050-000000000000 2513`)
}
benchM = m
}
func Benchmark_ParseLine_InfluxLog_NoMatch(b *testing.B) {
p := &Parser{
Patterns: []string{"%{INFLUXDB_HTTPD_LOG}"},
}
p.Compile()
var m telegraf.Metric
for n := 0; n < b.N; n++ {
m, _ = p.ParseLine(`[retention] 2016/06/14 14:38:24 retention policy shard deletion check commencing`)
}
benchM = m
}
func Benchmark_ParseLine_CustomPattern(b *testing.B) {
p := &Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatterns: `
DURATION %{NUMBER}[nuµm]?s
RESPONSE_CODE %{NUMBER:response_code:tag}
RESPONSE_TIME %{DURATION:response_time:duration}
TEST_LOG_A %{NUMBER:myfloat:float} %{RESPONSE_CODE} %{IPORHOST:clientip} %{RESPONSE_TIME}
`,
}
p.Compile()
var m telegraf.Metric
for n := 0; n < b.N; n++ {
m, _ = p.ParseLine(`[04/Jun/2016:12:41:45 +0100] 1.25 200 192.168.1.1 5.432µs 101`)
}
benchM = m
}
func TestBuiltinInfluxdbHttpd(t *testing.T) {
p := &Parser{
Patterns: []string{"%{INFLUXDB_HTTPD_LOG}"},
}
assert.NoError(t, p.Compile())
// Parse an influxdb POST request
m, err := p.ParseLine(`[httpd] ::1 - - [14/Jun/2016:11:33:29 +0100] "POST /write?consistency=any&db=telegraf&precision=ns&rp= HTTP/1.1" 204 0 "-" "InfluxDBClient" 6f61bc44-321b-11e6-8050-000000000000 2513`)
require.NotNil(t, m)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"resp_bytes": int64(0),
"auth": "-",
"client_ip": "::1",
"resp_code": int64(204),
"http_version": float64(1.1),
"ident": "-",
"referrer": "-",
"request": "/write?consistency=any&db=telegraf&precision=ns&rp=",
"response_time_us": int64(2513),
"agent": "InfluxDBClient",
},
m.Fields())
assert.Equal(t, map[string]string{"verb": "POST"}, m.Tags())
// Parse an influxdb GET request
m, err = p.ParseLine(`[httpd] ::1 - - [14/Jun/2016:12:10:02 +0100] "GET /query?db=telegraf&q=SELECT+bytes%2Cresponse_time_us+FROM+logparser_grok+WHERE+http_method+%3D+%27GET%27+AND+response_time_us+%3E+0+AND+time+%3E+now%28%29+-+1h HTTP/1.1" 200 578 "http://localhost:8083/" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.84 Safari/537.36" 8a3806f1-3220-11e6-8006-000000000000 988`)
require.NotNil(t, m)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"resp_bytes": int64(578),
"auth": "-",
"client_ip": "::1",
"resp_code": int64(200),
"http_version": float64(1.1),
"ident": "-",
"referrer": "http://localhost:8083/",
"request": "/query?db=telegraf&q=SELECT+bytes%2Cresponse_time_us+FROM+logparser_grok+WHERE+http_method+%3D+%27GET%27+AND+response_time_us+%3E+0+AND+time+%3E+now%28%29+-+1h",
"response_time_us": int64(988),
"agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.84 Safari/537.36",
},
m.Fields())
assert.Equal(t, map[string]string{"verb": "GET"}, m.Tags())
}
// common log format
// 127.0.0.1 user-identifier frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326
func TestBuiltinCommonLogFormat(t *testing.T) {
p := &Parser{
Patterns: []string{"%{COMMON_LOG_FORMAT}"},
}
assert.NoError(t, p.Compile())
// Parse an influxdb POST request
m, err := p.ParseLine(`127.0.0.1 user-identifier frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326`)
require.NotNil(t, m)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"resp_bytes": int64(2326),
"auth": "frank",
"client_ip": "127.0.0.1",
"resp_code": int64(200),
"http_version": float64(1.0),
"ident": "user-identifier",
"request": "/apache_pb.gif",
},
m.Fields())
assert.Equal(t, map[string]string{"verb": "GET"}, m.Tags())
}
// combined log format
// 127.0.0.1 user-identifier frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326 "-" "Mozilla"
func TestBuiltinCombinedLogFormat(t *testing.T) {
p := &Parser{
Patterns: []string{"%{COMBINED_LOG_FORMAT}"},
}
assert.NoError(t, p.Compile())
// Parse an influxdb POST request
m, err := p.ParseLine(`127.0.0.1 user-identifier frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326 "-" "Mozilla"`)
require.NotNil(t, m)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"resp_bytes": int64(2326),
"auth": "frank",
"client_ip": "127.0.0.1",
"resp_code": int64(200),
"http_version": float64(1.0),
"ident": "user-identifier",
"request": "/apache_pb.gif",
"referrer": "-",
"agent": "Mozilla",
},
m.Fields())
assert.Equal(t, map[string]string{"verb": "GET"}, m.Tags())
}
func TestCompileStringAndParse(t *testing.T) {
p := &Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatterns: `
DURATION %{NUMBER}[nuµm]?s
RESPONSE_CODE %{NUMBER:response_code:tag}
RESPONSE_TIME %{DURATION:response_time:duration}
TEST_LOG_A %{NUMBER:myfloat:float} %{RESPONSE_CODE} %{IPORHOST:clientip} %{RESPONSE_TIME}
`,
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`1.25 200 192.168.1.1 5.432µs`)
require.NotNil(t, metricA)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"clientip": "192.168.1.1",
"myfloat": float64(1.25),
"response_time": int64(5432),
},
metricA.Fields())
assert.Equal(t, map[string]string{"response_code": "200"}, metricA.Tags())
}
func TestParseEpochNano(t *testing.T) {
p := &Parser{
Patterns: []string{"%{MYAPP}"},
CustomPatterns: `
MYAPP %{POSINT:ts:ts-epochnano} response_time=%{POSINT:response_time:int} mymetric=%{NUMBER:metric:float}
`,
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`1466004605359052000 response_time=20821 mymetric=10890.645`)
require.NotNil(t, metricA)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"response_time": int64(20821),
"metric": float64(10890.645),
},
metricA.Fields())
assert.Equal(t, map[string]string{}, metricA.Tags())
assert.Equal(t, time.Unix(0, 1466004605359052000), metricA.Time())
}
func TestParseEpoch(t *testing.T) {
p := &Parser{
Patterns: []string{"%{MYAPP}"},
CustomPatterns: `
MYAPP %{POSINT:ts:ts-epoch} response_time=%{POSINT:response_time:int} mymetric=%{NUMBER:metric:float}
`,
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`1466004605 response_time=20821 mymetric=10890.645`)
require.NotNil(t, metricA)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"response_time": int64(20821),
"metric": float64(10890.645),
},
metricA.Fields())
assert.Equal(t, map[string]string{}, metricA.Tags())
assert.Equal(t, time.Unix(1466004605, 0), metricA.Time())
}
func TestParseEpochErrors(t *testing.T) {
p := &Parser{
Patterns: []string{"%{MYAPP}"},
CustomPatterns: `
MYAPP %{WORD:ts:ts-epoch} response_time=%{POSINT:response_time:int} mymetric=%{NUMBER:metric:float}
`,
}
assert.NoError(t, p.Compile())
_, err := p.ParseLine(`foobar response_time=20821 mymetric=10890.645`)
assert.NoError(t, err)
p = &Parser{
Patterns: []string{"%{MYAPP}"},
CustomPatterns: `
MYAPP %{WORD:ts:ts-epochnano} response_time=%{POSINT:response_time:int} mymetric=%{NUMBER:metric:float}
`,
}
assert.NoError(t, p.Compile())
_, err = p.ParseLine(`foobar response_time=20821 mymetric=10890.645`)
assert.NoError(t, err)
}
func TestCompileFileAndParse(t *testing.T) {
p := &Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatternFiles: []string{"./testdata/test-patterns"},
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`[04/Jun/2016:12:41:45 +0100] 1.25 200 192.168.1.1 5.432µs 101`)
require.NotNil(t, metricA)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"clientip": "192.168.1.1",
"myfloat": float64(1.25),
"response_time": int64(5432),
"myint": int64(101),
},
metricA.Fields())
assert.Equal(t, map[string]string{"response_code": "200"}, metricA.Tags())
assert.Equal(t,
time.Date(2016, time.June, 4, 12, 41, 45, 0, time.FixedZone("foo", 60*60)).Nanosecond(),
metricA.Time().Nanosecond())
metricB, err := p.ParseLine(`[04/06/2016--12:41:45] 1.25 mystring dropme nomodifier`)
require.NotNil(t, metricB)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"myfloat": 1.25,
"mystring": "mystring",
"nomodifier": "nomodifier",
},
metricB.Fields())
assert.Equal(t, map[string]string{}, metricB.Tags())
assert.Equal(t,
time.Date(2016, time.June, 4, 12, 41, 45, 0, time.FixedZone("foo", 60*60)).Nanosecond(),
metricB.Time().Nanosecond())
}
func TestCompileNoModifiersAndParse(t *testing.T) {
p := &Parser{
Patterns: []string{"%{TEST_LOG_C}"},
CustomPatterns: `
DURATION %{NUMBER}[nuµm]?s
TEST_LOG_C %{NUMBER:myfloat} %{NUMBER} %{IPORHOST:clientip} %{DURATION:rt}
`,
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`1.25 200 192.168.1.1 5.432µs`)
require.NotNil(t, metricA)
assert.NoError(t, err)
assert.Equal(t,
map[string]interface{}{
"clientip": "192.168.1.1",
"myfloat": "1.25",
"rt": "5.432µs",
},
metricA.Fields())
assert.Equal(t, map[string]string{}, metricA.Tags())
}
func TestCompileNoNamesAndParse(t *testing.T) {
p := &Parser{
Patterns: []string{"%{TEST_LOG_C}"},
CustomPatterns: `
DURATION %{NUMBER}[nuµm]?s
TEST_LOG_C %{NUMBER} %{NUMBER} %{IPORHOST} %{DURATION}
`,
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`1.25 200 192.168.1.1 5.432µs`)
require.Nil(t, metricA)
assert.NoError(t, err)
}
func TestParseNoMatch(t *testing.T) {
p := &Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatternFiles: []string{"./testdata/test-patterns"},
}
assert.NoError(t, p.Compile())
metricA, err := p.ParseLine(`[04/Jun/2016:12:41:45 +0100] notnumber 200 192.168.1.1 5.432µs 101`)
assert.NoError(t, err)
assert.Nil(t, metricA)
}
func TestCompileErrors(t *testing.T) {
// Compile fails because there are multiple timestamps:
p := &Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatterns: `
TEST_LOG_A %{HTTPDATE:ts1:ts-httpd} %{HTTPDATE:ts2:ts-httpd} %{NUMBER:mynum:int}
`,
}
assert.Error(t, p.Compile())
// Compile fails because file doesn't exist:
p = &Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatternFiles: []string{"/tmp/foo/bar/baz"},
}
assert.Error(t, p.Compile())
}
func TestParseErrors(t *testing.T) {
// Parse fails because the pattern doesn't exist
p := &Parser{
Patterns: []string{"%{TEST_LOG_B}"},
CustomPatterns: `
TEST_LOG_A %{HTTPDATE:ts:ts-httpd} %{WORD:myword:int} %{}
`,
}
assert.NoError(t, p.Compile())
_, err := p.ParseLine(`[04/Jun/2016:12:41:45 +0100] notnumber 200 192.168.1.1 5.432µs 101`)
assert.Error(t, err)
// Parse fails because myword is not an int
p = &Parser{
Patterns: []string{"%{TEST_LOG_A}"},
CustomPatterns: `
TEST_LOG_A %{HTTPDATE:ts:ts-httpd} %{WORD:myword:int}
`,
}
assert.NoError(t, p.Compile())
_, err = p.ParseLine(`04/Jun/2016:12:41:45 +0100 notnumber`)
assert.Error(t, err)
// Parse fails because myword is not a float
p = &Parser{
Patterns: []string{"%{TEST_LOG_A}"},
CustomPatterns: `
TEST_LOG_A %{HTTPDATE:ts:ts-httpd} %{WORD:myword:float}
`,
}
assert.NoError(t, p.Compile())
_, err = p.ParseLine(`04/Jun/2016:12:41:45 +0100 notnumber`)
assert.Error(t, err)
// Parse fails because myword is not a duration
p = &Parser{
Patterns: []string{"%{TEST_LOG_A}"},
CustomPatterns: `
TEST_LOG_A %{HTTPDATE:ts:ts-httpd} %{WORD:myword:duration}
`,
}
assert.NoError(t, p.Compile())
_, err = p.ParseLine(`04/Jun/2016:12:41:45 +0100 notnumber`)
assert.Error(t, err)
// Parse fails because the time layout is wrong.
p = &Parser{
Patterns: []string{"%{TEST_LOG_A}"},
CustomPatterns: `
TEST_LOG_A %{HTTPDATE:ts:ts-unix} %{WORD:myword:duration}
`,
}
assert.NoError(t, p.Compile())
_, err = p.ParseLine(`04/Jun/2016:12:41:45 +0100 notnumber`)
assert.Error(t, err)
}
func TestTsModder(t *testing.T) {
tsm := &tsModder{}
reftime := time.Date(2006, time.December, 1, 1, 1, 1, int(time.Millisecond), time.UTC)
modt := tsm.tsMod(reftime)
assert.Equal(t, reftime, modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Microsecond*1), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Microsecond*2), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Microsecond*3), modt)
reftime = time.Date(2006, time.December, 1, 1, 1, 1, int(time.Microsecond), time.UTC)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime, modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Nanosecond*1), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Nanosecond*2), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Nanosecond*3), modt)
reftime = time.Date(2006, time.December, 1, 1, 1, 1, int(time.Microsecond)*999, time.UTC)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime, modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Nanosecond*1), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Nanosecond*2), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Nanosecond*3), modt)
reftime = time.Date(2006, time.December, 1, 1, 1, 1, 0, time.UTC)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime, modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Millisecond*1), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Millisecond*2), modt)
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime.Add(time.Millisecond*3), modt)
reftime = time.Time{}
modt = tsm.tsMod(reftime)
assert.Equal(t, reftime, modt)
}
func TestTsModder_Rollover(t *testing.T) {
tsm := &tsModder{}
reftime := time.Date(2006, time.December, 1, 1, 1, 1, int(time.Millisecond), time.UTC)
modt := tsm.tsMod(reftime)
for i := 1; i < 1000; i++ {
modt = tsm.tsMod(reftime)
}
assert.Equal(t, reftime.Add(time.Microsecond*999+time.Nanosecond), modt)
reftime = time.Date(2006, time.December, 1, 1, 1, 1, int(time.Microsecond), time.UTC)
modt = tsm.tsMod(reftime)
for i := 1; i < 1001; i++ {
modt = tsm.tsMod(reftime)
}
assert.Equal(t, reftime.Add(time.Nanosecond*1000), modt)
}

View File

@ -0,0 +1,80 @@
package grok
// THIS SHOULD BE KEPT IN-SYNC WITH patterns/influx-patterns
const DEFAULT_PATTERNS = `
# Captures are a slightly modified version of logstash "grok" patterns, with
# the format %{<capture syntax>[:<semantic name>][:<modifier>]}
# By default all named captures are converted into string fields.
# Modifiers can be used to convert captures to other types or tags.
# Timestamp modifiers can be used to convert captures to the timestamp of the
# parsed metric.
# View logstash grok pattern docs here:
# https://www.elastic.co/guide/en/logstash/current/plugins-filters-grok.html
# All default logstash patterns are supported, these can be viewed here:
# https://github.com/logstash-plugins/logstash-patterns-core/blob/master/patterns/grok-patterns
# Available modifiers:
# string (default if nothing is specified)
# int
# float
# duration (ie, 5.23ms gets converted to int nanoseconds)
# tag (converts the field into a tag)
# drop (drops the field completely)
# Timestamp modifiers:
# ts-ansic ("Mon Jan _2 15:04:05 2006")
# ts-unix ("Mon Jan _2 15:04:05 MST 2006")
# ts-ruby ("Mon Jan 02 15:04:05 -0700 2006")
# ts-rfc822 ("02 Jan 06 15:04 MST")
# ts-rfc822z ("02 Jan 06 15:04 -0700")
# ts-rfc850 ("Monday, 02-Jan-06 15:04:05 MST")
# ts-rfc1123 ("Mon, 02 Jan 2006 15:04:05 MST")
# ts-rfc1123z ("Mon, 02 Jan 2006 15:04:05 -0700")
# ts-rfc3339 ("2006-01-02T15:04:05Z07:00")
# ts-rfc3339nano ("2006-01-02T15:04:05.999999999Z07:00")
# ts-httpd ("02/Jan/2006:15:04:05 -0700")
# ts-epoch (seconds since unix epoch)
# ts-epochnano (nanoseconds since unix epoch)
# ts-"CUSTOM"
# CUSTOM time layouts must be within quotes and be the representation of the
# "reference time", which is Mon Jan 2 15:04:05 -0700 MST 2006
# See https://golang.org/pkg/time/#Parse for more details.
# Example log file pattern, example log looks like this:
# [04/Jun/2016:12:41:45 +0100] 1.25 200 192.168.1.1 5.432µs
# Breakdown of the DURATION pattern below:
# NUMBER is a builtin logstash grok pattern matching float & int numbers.
# [nuµm]? is a regex specifying 0 or 1 of the characters within brackets.
# s is also regex, this pattern must end in "s".
# so DURATION will match something like '5.324ms' or '6.1µs' or '10s'
DURATION %{NUMBER}[nuµm]?s
RESPONSE_CODE %{NUMBER:response_code:tag}
RESPONSE_TIME %{DURATION:response_time_ns:duration}
EXAMPLE_LOG \[%{HTTPDATE:ts:ts-httpd}\] %{NUMBER:myfloat:float} %{RESPONSE_CODE} %{IPORHOST:clientip} %{RESPONSE_TIME}
# Wider-ranging username matching vs. logstash built-in %{USER}
NGUSERNAME [a-zA-Z\.\@\-\+_%]+
NGUSER %{NGUSERNAME}
##
## COMMON LOG PATTERNS
##
# InfluxDB log patterns
CLIENT (?:%{IPORHOST}|%{HOSTPORT}|::1)
INFLUXDB_HTTPD_LOG \[httpd\] %{COMBINED_LOG_FORMAT} %{UUID:uuid:drop} %{NUMBER:response_time_us:int}
# apache & nginx logs, this is also known as the "common log format"
# see https://en.wikipedia.org/wiki/Common_Log_Format
COMMON_LOG_FORMAT %{CLIENT:client_ip} %{NGUSER:ident} %{NGUSER:auth} \[%{HTTPDATE:ts:ts-httpd}\] "(?:%{WORD:verb:tag} %{NOTSPACE:request}(?: HTTP/%{NUMBER:http_version:float})?|%{DATA})" %{NUMBER:resp_code:int} (?:%{NUMBER:resp_bytes:int}|-)
# Combined log format is the same as the common log format but with the addition
# of two quoted strings at the end for "referrer" and "agent"
# See Examples at http://httpd.apache.org/docs/current/mod/mod_log_config.html
COMBINED_LOG_FORMAT %{COMMON_LOG_FORMAT} %{QS:referrer} %{QS:agent}
# HTTPD log formats
HTTPD20_ERRORLOG \[%{HTTPDERROR_DATE:timestamp}\] \[%{LOGLEVEL:loglevel:tag}\] (?:\[client %{IPORHOST:clientip}\] ){0,1}%{GREEDYDATA:errormsg}
HTTPD24_ERRORLOG \[%{HTTPDERROR_DATE:timestamp}\] \[%{WORD:module}:%{LOGLEVEL:loglevel:tag}\] \[pid %{POSINT:pid:int}:tid %{NUMBER:tid:int}\]( \(%{POSINT:proxy_errorcode:int}\)%{DATA:proxy_errormessage}:)?( \[client %{IPORHOST:client}:%{POSINT:clientport}\])? %{DATA:errorcode}: %{GREEDYDATA:message}
HTTPD_ERRORLOG %{HTTPD20_ERRORLOG}|%{HTTPD24_ERRORLOG}
`

View File

@ -0,0 +1,75 @@
# Captures are a slightly modified version of logstash "grok" patterns, with
# the format %{<capture syntax>[:<semantic name>][:<modifier>]}
# By default all named captures are converted into string fields.
# Modifiers can be used to convert captures to other types or tags.
# Timestamp modifiers can be used to convert captures to the timestamp of the
# parsed metric.
# View logstash grok pattern docs here:
# https://www.elastic.co/guide/en/logstash/current/plugins-filters-grok.html
# All default logstash patterns are supported, these can be viewed here:
# https://github.com/logstash-plugins/logstash-patterns-core/blob/master/patterns/grok-patterns
# Available modifiers:
# string (default if nothing is specified)
# int
# float
# duration (ie, 5.23ms gets converted to int nanoseconds)
# tag (converts the field into a tag)
# drop (drops the field completely)
# Timestamp modifiers:
# ts-ansic ("Mon Jan _2 15:04:05 2006")
# ts-unix ("Mon Jan _2 15:04:05 MST 2006")
# ts-ruby ("Mon Jan 02 15:04:05 -0700 2006")
# ts-rfc822 ("02 Jan 06 15:04 MST")
# ts-rfc822z ("02 Jan 06 15:04 -0700")
# ts-rfc850 ("Monday, 02-Jan-06 15:04:05 MST")
# ts-rfc1123 ("Mon, 02 Jan 2006 15:04:05 MST")
# ts-rfc1123z ("Mon, 02 Jan 2006 15:04:05 -0700")
# ts-rfc3339 ("2006-01-02T15:04:05Z07:00")
# ts-rfc3339nano ("2006-01-02T15:04:05.999999999Z07:00")
# ts-httpd ("02/Jan/2006:15:04:05 -0700")
# ts-epoch (seconds since unix epoch)
# ts-epochnano (nanoseconds since unix epoch)
# ts-"CUSTOM"
# CUSTOM time layouts must be within quotes and be the representation of the
# "reference time", which is Mon Jan 2 15:04:05 -0700 MST 2006
# See https://golang.org/pkg/time/#Parse for more details.
# Example log file pattern, example log looks like this:
# [04/Jun/2016:12:41:45 +0100] 1.25 200 192.168.1.1 5.432µs
# Breakdown of the DURATION pattern below:
# NUMBER is a builtin logstash grok pattern matching float & int numbers.
# [nuµm]? is a regex specifying 0 or 1 of the characters within brackets.
# s is also regex, this pattern must end in "s".
# so DURATION will match something like '5.324ms' or '6.1µs' or '10s'
DURATION %{NUMBER}[nuµm]?s
RESPONSE_CODE %{NUMBER:response_code:tag}
RESPONSE_TIME %{DURATION:response_time_ns:duration}
EXAMPLE_LOG \[%{HTTPDATE:ts:ts-httpd}\] %{NUMBER:myfloat:float} %{RESPONSE_CODE} %{IPORHOST:clientip} %{RESPONSE_TIME}
# Wider-ranging username matching vs. logstash built-in %{USER}
NGUSERNAME [a-zA-Z\.\@\-\+_%]+
NGUSER %{NGUSERNAME}
##
## COMMON LOG PATTERNS
##
# InfluxDB log patterns
CLIENT (?:%{IPORHOST}|%{HOSTPORT}|::1)
INFLUXDB_HTTPD_LOG \[httpd\] %{COMBINED_LOG_FORMAT} %{UUID:uuid:drop} %{NUMBER:response_time_us:int}
# apache & nginx logs, this is also known as the "common log format"
# see https://en.wikipedia.org/wiki/Common_Log_Format
COMMON_LOG_FORMAT %{CLIENT:client_ip} %{NGUSER:ident} %{NGUSER:auth} \[%{HTTPDATE:ts:ts-httpd}\] "(?:%{WORD:verb:tag} %{NOTSPACE:request}(?: HTTP/%{NUMBER:http_version:float})?|%{DATA})" %{NUMBER:resp_code:int} (?:%{NUMBER:resp_bytes:int}|-)
# Combined log format is the same as the common log format but with the addition
# of two quoted strings at the end for "referrer" and "agent"
# See Examples at http://httpd.apache.org/docs/current/mod/mod_log_config.html
COMBINED_LOG_FORMAT %{COMMON_LOG_FORMAT} %{QS:referrer} %{QS:agent}
# HTTPD log formats
HTTPD20_ERRORLOG \[%{HTTPDERROR_DATE:timestamp}\] \[%{LOGLEVEL:loglevel:tag}\] (?:\[client %{IPORHOST:clientip}\] ){0,1}%{GREEDYDATA:errormsg}
HTTPD24_ERRORLOG \[%{HTTPDERROR_DATE:timestamp}\] \[%{WORD:module}:%{LOGLEVEL:loglevel:tag}\] \[pid %{POSINT:pid:int}:tid %{NUMBER:tid:int}\]( \(%{POSINT:proxy_errorcode:int}\)%{DATA:proxy_errormessage}:)?( \[client %{IPORHOST:client}:%{POSINT:clientport}\])? %{DATA:errorcode}: %{GREEDYDATA:message}
HTTPD_ERRORLOG %{HTTPD20_ERRORLOG}|%{HTTPD24_ERRORLOG}

View File

@ -0,0 +1,14 @@
# Test A log line:
# [04/Jun/2016:12:41:45 +0100] 1.25 200 192.168.1.1 5.432µs 101
DURATION %{NUMBER}[nuµm]?s
RESPONSE_CODE %{NUMBER:response_code:tag}
RESPONSE_TIME %{DURATION:response_time:duration}
TEST_LOG_A \[%{HTTPDATE:timestamp:ts-httpd}\] %{NUMBER:myfloat:float} %{RESPONSE_CODE} %{IPORHOST:clientip} %{RESPONSE_TIME} %{NUMBER:myint:int}
# Test B log line:
# [04/06/2016--12:41:45] 1.25 mystring dropme nomodifier
TEST_TIMESTAMP %{MONTHDAY}/%{MONTHNUM}/%{YEAR}--%{TIME}
TEST_LOG_B \[%{TEST_TIMESTAMP:timestamp:ts-"02/01/2006--15:04:05"}\] %{NUMBER:myfloat:float} %{WORD:mystring:string} %{WORD:dropme:drop} %{WORD:nomodifier}
TEST_TIMESTAMP %{MONTHDAY}/%{MONTHNUM}/%{YEAR}--%{TIME}
TEST_LOG_BAD \[%{TEST_TIMESTAMP:timestamp:ts-"02/01/2006--15:04:05"}\] %{NUMBER:myfloat:float} %{WORD:mystring:int} %{WORD:dropme:drop} %{WORD:nomodifier}

View File

@ -0,0 +1 @@
[04/Jun/2016:12:41:45 +0100] 1.25 200 192.168.1.1 5.432µs 101

View File

@ -0,0 +1 @@
[04/06/2016--12:41:45] 1.25 mystring dropme nomodifier

View File

@ -0,0 +1,228 @@
package logparser
import (
"fmt"
"log"
"reflect"
"sync"
"github.com/hpcloud/tail"
"github.com/influxdata/telegraf"
"github.com/influxdata/telegraf/internal/globpath"
"github.com/influxdata/telegraf/plugins/inputs"
// Parsers
"github.com/influxdata/telegraf/plugins/inputs/logparser/grok"
)
type LogParser interface {
ParseLine(line string) (telegraf.Metric, error)
Compile() error
}
type LogParserPlugin struct {
Files []string
FromBeginning bool
tailers []*tail.Tail
lines chan string
done chan struct{}
wg sync.WaitGroup
acc telegraf.Accumulator
parsers []LogParser
sync.Mutex
GrokParser *grok.Parser `toml:"grok"`
}
const sampleConfig = `
## Log files to parse.
## These accept standard unix glob matching rules, but with the addition of
## ** as a "super asterisk". ie:
## /var/log/**.log -> recursively find all .log files in /var/log
## /var/log/*/*.log -> find all .log files with a parent dir in /var/log
## /var/log/apache.log -> only tail the apache log file
files = ["/var/log/influxdb/influxdb.log"]
## Read file from beginning.
from_beginning = false
## Parse logstash-style "grok" patterns:
## Telegraf built-in parsing patterns: https://goo.gl/dkay10
[inputs.logparser.grok]
## This is a list of patterns to check the given log file(s) for.
## Note that adding patterns here increases processing time. The most
## efficient configuration is to have one pattern per logparser.
## Other common built-in patterns are:
## %{COMMON_LOG_FORMAT} (plain apache & nginx access logs)
## %{COMBINED_LOG_FORMAT} (access logs + referrer & agent)
patterns = ["%{INFLUXDB_HTTPD_LOG}"]
## Full path(s) to custom pattern files.
custom_pattern_files = []
## Custom patterns can also be defined here. Put one pattern per line.
custom_patterns = '''
'''
`
func (l *LogParserPlugin) SampleConfig() string {
return sampleConfig
}
func (l *LogParserPlugin) Description() string {
return "Stream and parse log file(s)."
}
func (l *LogParserPlugin) Gather(acc telegraf.Accumulator) error {
return nil
}
func (l *LogParserPlugin) Start(acc telegraf.Accumulator) error {
l.Lock()
defer l.Unlock()
l.acc = acc
l.lines = make(chan string, 1000)
l.done = make(chan struct{})
// Looks for fields which implement LogParser interface
l.parsers = []LogParser{}
s := reflect.ValueOf(l).Elem()
for i := 0; i < s.NumField(); i++ {
f := s.Field(i)
if !f.CanInterface() {
continue
}
if lpPlugin, ok := f.Interface().(LogParser); ok {
if reflect.ValueOf(lpPlugin).IsNil() {
continue
}
l.parsers = append(l.parsers, lpPlugin)
}
}
if len(l.parsers) == 0 {
return fmt.Errorf("ERROR: logparser input plugin: no parser defined.")
}
// compile log parser patterns:
for _, parser := range l.parsers {
if err := parser.Compile(); err != nil {
return err
}
}
var seek tail.SeekInfo
if !l.FromBeginning {
seek.Whence = 2
seek.Offset = 0
}
l.wg.Add(1)
go l.parser()
var errS string
// Create a "tailer" for each file
for _, filepath := range l.Files {
g, err := globpath.Compile(filepath)
if err != nil {
log.Printf("ERROR Glob %s failed to compile, %s", filepath, err)
}
for file, _ := range g.Match() {
tailer, err := tail.TailFile(file,
tail.Config{
ReOpen: true,
Follow: true,
Location: &seek,
})
if err != nil {
errS += err.Error() + " "
continue
}
// create a goroutine for each "tailer"
l.wg.Add(1)
go l.receiver(tailer)
l.tailers = append(l.tailers, tailer)
}
}
if errS != "" {
return fmt.Errorf(errS)
}
return nil
}
// receiver is launched as a goroutine to continuously watch a tailed logfile
// for changes and send any log lines down the l.lines channel.
func (l *LogParserPlugin) receiver(tailer *tail.Tail) {
defer l.wg.Done()
var line *tail.Line
for line = range tailer.Lines {
if line.Err != nil {
log.Printf("ERROR tailing file %s, Error: %s\n",
tailer.Filename, line.Err)
continue
}
select {
case <-l.done:
case l.lines <- line.Text:
}
}
}
// parser is launched as a goroutine to watch the l.lines channel.
// when a line is available, parser parses it and adds the metric(s) to the
// accumulator.
func (l *LogParserPlugin) parser() {
defer l.wg.Done()
var m telegraf.Metric
var err error
var line string
for {
select {
case <-l.done:
return
case line = <-l.lines:
if line == "" || line == "\n" {
continue
}
}
for _, parser := range l.parsers {
m, err = parser.ParseLine(line)
if err == nil {
if m != nil {
l.acc.AddFields(m.Name(), m.Fields(), m.Tags(), m.Time())
}
} else {
log.Printf("Malformed log line in [%s], Error: %s\n", line, err)
}
}
}
}
func (l *LogParserPlugin) Stop() {
l.Lock()
defer l.Unlock()
for _, t := range l.tailers {
err := t.Stop()
if err != nil {
log.Printf("ERROR stopping tail on file %s\n", t.Filename)
}
t.Cleanup()
}
close(l.done)
l.wg.Wait()
}
func init() {
inputs.Add("logparser", func() telegraf.Input {
return &LogParserPlugin{}
})
}

View File

@ -0,0 +1,116 @@
package logparser
import (
"runtime"
"strings"
"testing"
"time"
"github.com/influxdata/telegraf/testutil"
"github.com/influxdata/telegraf/plugins/inputs/logparser/grok"
"github.com/stretchr/testify/assert"
)
func TestStartNoParsers(t *testing.T) {
logparser := &LogParserPlugin{
FromBeginning: true,
Files: []string{"grok/testdata/*.log"},
}
acc := testutil.Accumulator{}
assert.Error(t, logparser.Start(&acc))
}
func TestGrokParseLogFilesNonExistPattern(t *testing.T) {
thisdir := getCurrentDir()
p := &grok.Parser{
Patterns: []string{"%{FOOBAR}"},
CustomPatternFiles: []string{thisdir + "grok/testdata/test-patterns"},
}
logparser := &LogParserPlugin{
FromBeginning: true,
Files: []string{thisdir + "grok/testdata/*.log"},
GrokParser: p,
}
acc := testutil.Accumulator{}
assert.NoError(t, logparser.Start(&acc))
time.Sleep(time.Millisecond * 500)
logparser.Stop()
}
func TestGrokParseLogFiles(t *testing.T) {
thisdir := getCurrentDir()
p := &grok.Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_B}"},
CustomPatternFiles: []string{thisdir + "grok/testdata/test-patterns"},
}
logparser := &LogParserPlugin{
FromBeginning: true,
Files: []string{thisdir + "grok/testdata/*.log"},
GrokParser: p,
}
acc := testutil.Accumulator{}
assert.NoError(t, logparser.Start(&acc))
time.Sleep(time.Millisecond * 500)
logparser.Stop()
acc.AssertContainsTaggedFields(t, "logparser_grok",
map[string]interface{}{
"clientip": "192.168.1.1",
"myfloat": float64(1.25),
"response_time": int64(5432),
"myint": int64(101),
},
map[string]string{"response_code": "200"})
acc.AssertContainsTaggedFields(t, "logparser_grok",
map[string]interface{}{
"myfloat": 1.25,
"mystring": "mystring",
"nomodifier": "nomodifier",
},
map[string]string{})
}
func TestGrokParseLogFilesOneBad(t *testing.T) {
thisdir := getCurrentDir()
p := &grok.Parser{
Patterns: []string{"%{TEST_LOG_A}", "%{TEST_LOG_BAD}"},
CustomPatternFiles: []string{thisdir + "grok/testdata/test-patterns"},
}
assert.NoError(t, p.Compile())
logparser := &LogParserPlugin{
FromBeginning: true,
Files: []string{thisdir + "grok/testdata/*.log"},
GrokParser: p,
}
acc := testutil.Accumulator{}
assert.NoError(t, logparser.Start(&acc))
time.Sleep(time.Millisecond * 500)
logparser.Stop()
acc.AssertContainsTaggedFields(t, "logparser_grok",
map[string]interface{}{
"clientip": "192.168.1.1",
"myfloat": float64(1.25),
"response_time": int64(5432),
"myint": int64(101),
},
map[string]string{"response_code": "200"})
}
func getCurrentDir() string {
_, filename, _, _ := runtime.Caller(1)
return strings.Replace(filename, "logparser_test.go", "", 1)
}

View File

@ -12,9 +12,8 @@ import (
"strings"
"time"
"github.com/gobwas/glob"
"github.com/influxdata/telegraf"
"github.com/influxdata/telegraf/filter"
"github.com/influxdata/telegraf/internal"
"github.com/influxdata/telegraf/plugins/inputs"
)
@ -26,7 +25,7 @@ type Varnish struct {
Stats []string
Binary string
filter glob.Glob
filter filter.Filter
run runner
}
@ -78,13 +77,13 @@ func (s *Varnish) Gather(acc telegraf.Accumulator) error {
if s.filter == nil {
var err error
if len(s.Stats) == 0 {
s.filter, err = internal.CompileFilter(defaultStats)
s.filter, err = filter.CompileFilter(defaultStats)
} else {
// legacy support, change "all" -> "*":
if s.Stats[0] == "all" {
s.Stats[0] = "*"
}
s.filter, err = internal.CompileFilter(s.Stats)
s.filter, err = filter.CompileFilter(s.Stats)
}
if err != nil {
return err