telegraf/plugins/parsers/grok
Daniel Nelson b6cc324d10
Add note on performance to grok parser documentation (#5291)
2019-01-17 12:00:07 -08:00
..
testdata Remove DS_Store files 2018-07-13 23:23:47 -07:00
README.md Add note on performance to grok parser documentation (#5291) 2019-01-17 12:00:07 -08:00
influx_patterns.go Improve docs to clarify common issues (#5054) 2018-12-10 16:14:20 -08:00
parser.go Fix divide by zero in logparser input (#4338) 2018-08-29 19:11:13 -07:00
parser_test.go Fix unittests for new year (#5213) 2019-01-02 11:12:04 -08:00

README.md

Grok

The grok data format parses line delimited data using a regular expression like language.

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

The grok parser uses a slightly modified version of logstash "grok" patterns, with the format:

%{<capture_syntax>[:<semantic_name>][:<modifier>]}

The capture_syntax defines the grok pattern that's used to parse the input line and the semantic_name is used to name the field or tag. The extension modifier controls the data type that the parsed item is converted to or other special handling.

By default all named captures are converted into string fields. If a pattern does not have a semantic name it will not be captured. Timestamp modifiers can be used to convert captures to the timestamp of the parsed metric. If no timestamp is parsed the metric will be created using the current time.

You must capture at least one field per line.

  • 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)
    • measurement (use the matched text as the measurement name)
  • Timestamp modifiers:
    • ts (This will auto-learn the timestamp format)
    • 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, may contain decimal)
    • ts-epochnano (nanoseconds since unix epoch)
    • ts-syslog ("Jan 02 15:04:05", parsed time is set to the current year)
    • 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. To match a comma decimal point you can use a period. For example %{TIMESTAMP:timestamp:ts-"2006-01-02 15:04:05.000"} can be used to match "2018-01-02 15:04:05,000" To match a comma decimal point you can use a period in the pattern string. See https://golang.org/pkg/time/#Parse for more details.

Telegraf has many of its own built-in patterns, as well as support for most of logstash's builtin patterns. Golang regular expressions do not support lookahead or lookbehind. logstash patterns that depend on these are not supported.

If you need help building patterns to match your logs, you will find the https://grokdebug.herokuapp.com application quite useful!

Configuration

[[inputs.file]]
  ## Files to parse each interval.
  ## 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/apache/access.log"]

  ## The dataformat to be read from files
  ## 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 = "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.
  ## Other common built-in patterns are:
  ##   %{COMMON_LOG_FORMAT}   (plain apache & nginx access logs)
  ##   %{COMBINED_LOG_FORMAT} (access logs + referrer & agent)
  grok_patterns = ["%{COMBINED_LOG_FORMAT}"]

  ## Full path(s) to custom pattern files.
  grok_custom_pattern_files = []

  ## Custom patterns can also be defined here. Put one pattern per line.
  grok_custom_patterns = '''
  '''

  ## Timezone allows you to provide an override for timestamps that
  ## don't already include an offset
  ## e.g. 04/06/2016 12:41:45 data one two 5.43µs
  ##
  ## Default: "" which renders UTC
  ## Options are as follows:
  ##   1. Local             -- interpret based on machine localtime
  ##   2. "Canada/Eastern"  -- Unix TZ values like those found in https://en.wikipedia.org/wiki/List_of_tz_database_time_zones
  ##   3. UTC               -- or blank/unspecified, will return timestamp in UTC
  grok_timezone = "Canada/Eastern"

Timestamp Examples

This example input and config parses a file using a custom timestamp conversion:

2017-02-21 13:10:34 value=42
[[inputs.file]]
  grok_patterns = ['%{TIMESTAMP_ISO8601:timestamp:ts-"2006-01-02 15:04:05"} value=%{NUMBER:value:int}']

This example input and config parses a file using a timestamp in unix time:

1466004605 value=42
1466004605.123456789 value=42
[[inputs.file]]
  grok_patterns = ['%{NUMBER:timestamp:ts-epoch} value=%{NUMBER:value:int}']

This example parses a file using a built-in conversion and a custom pattern:

Wed Apr 12 13:10:34 PST 2017 value=42
[[inputs.file]]
  grok_patterns = ["%{TS_UNIX:timestamp:ts-unix} value=%{NUMBER:value:int}"]
  grok_custom_patterns = '''
    TS_UNIX %{DAY} %{MONTH} %{MONTHDAY} %{HOUR}:%{MINUTE}:%{SECOND} %{TZ} %{YEAR}
  '''

For cases where the timestamp itself is without offset, the timezone config var is available to denote an offset. By default (with timezone either omit, blank or set to "UTC"), the times are processed as if in the UTC timezone. If specified as timezone = "Local", the timestamp will be processed based on the current machine timezone configuration. Lastly, if using a timezone from the list of Unix timezones, grok will offset the timestamp accordingly.

TOML Escaping

When saving patterns to the configuration file, keep in mind the different TOML string types and the escaping rules for each. These escaping rules must be applied in addition to the escaping required by the grok syntax. Using the Multi-line line literal syntax with ''' may be useful.

The following config examples will parse this input file:

|42|\uD83D\uDC2F|'telegraf'|

Since | is a special character in the grok language, we must escape it to get a literal |. With a basic TOML string, special characters such as backslash must be escaped, requiring us to escape the backslash a second time.

[[inputs.file]]
  grok_patterns = ["\\|%{NUMBER:value:int}\\|%{UNICODE_ESCAPE:escape}\\|'%{WORD:name}'\\|"]
  grok_custom_patterns = "UNICODE_ESCAPE (?:\\\\u[0-9A-F]{4})+"

We cannot use a literal TOML string for the pattern, because we cannot match a ' within it. However, it works well for the custom pattern.

[[inputs.file]]
  grok_patterns = ["\\|%{NUMBER:value:int}\\|%{UNICODE_ESCAPE:escape}\\|'%{WORD:name}'\\|"]
  grok_custom_patterns = 'UNICODE_ESCAPE (?:\\u[0-9A-F]{4})+'

A multi-line literal string allows us to encode the pattern:

[[inputs.file]]
  grok_patterns = ['''
    \|%{NUMBER:value:int}\|%{UNICODE_ESCAPE:escape}\|'%{WORD:name}'\|
  ''']
  grok_custom_patterns = 'UNICODE_ESCAPE (?:\\u[0-9A-F]{4})+'

Tips for creating patterns

Writing complex patterns can be difficult, here is some advice for writing a new pattern or testing a pattern developed online.

Create a file output that writes to stdout, and disable other outputs while testing. This will allow you to see the captured metrics. Keep in mind that the file output will only print once per flush_interval.

[[outputs.file]]
  files = ["stdout"]
  • Start with a file containing only a single line of your input.
  • Remove all but the first token or piece of the line.
  • Add the section of your pattern to match this piece to your configuration file.
  • Verify that the metric is parsed successfully by running Telegraf.
  • If successful, add the next token, update the pattern and retest.
  • Continue one token at a time until the entire line is successfully parsed.

Performance

Performance depends heavily on the regular expressions that you use, but there are a few techniques that can help:

  • Avoid using patterns such as %{DATA} that will always match.
  • If possible, add ^ and $ anchors to your pattern:
    [[inputs.file]]
      grok_patterns = ["^%{COMBINED_LOG_FORMAT}$"]