Search code examples
logginglogstashlogstash-grok

Grok parse data inside square brackets

What is the best way to parse this kind of log with grok?

2019-03-15 14:42:38,910 INFO  [SID:6cd1c3cc-7fb0-4a06-8d4d-e125382568ca] [CID:60c24e3e-c8f9-43e4-bedf-59e861bfabf9] [http-bio-8080-exec-3] [TariffRuleServiceImpl.java:569] no approach by ShouldAddApproachToFixedPrice checkbox : false

I would like to get this kind of result:

{
  "date": [
    "19-03-15"
  ],
  "time": [
    "14:42:38,910"
  ],
  "level": [
    "INFO"
  ],
 "SID": [
    "6cd1c3cc-7fb0-4a06-8d4d-e125382568ca"
  ],
 "CID": [
    "60c24e3e-c8f9-43e4-bedf-59e861bfabf9"
  ],
 "thread": [
    "http-bio-8080-exec-3"
  ],
 "class": [
    "TariffRuleServiceImpl.java:569"
  ],
 "message": [
    "no approach by ShouldAddApproachToFixedPrice checkbox : false"
  ]
}

But I am stuck here: %{DATE:date} %{TIME:time} %{WORD:level} ...? How to get data inside [...]? Thanks a lot

Solution

  • The main idea of your Grok pattern should be to skip the square brackets with a backslash as for example \[. Once you have skipped them you can use a regular pattern to match your elements. Something like this would work as you want:

    %{DATE:date} %{TIME:time} %{WORD:level}  \[SID:%{DATA:SID}\] \[CID:%{DATA:CID}\] \[%{DATA:thread}] \[%{DATA:class}] %{GREEDYDATA:message}

    Also, realize that you have two spaces between the level and the [SID and this could be making your pattern not to match depending on your construction.