Splunk Search

Extract values from JSON array

DimkoBilanko
Explorer

Hi everyone!
I have a JSON output in raw format:

{"result":{"addr":"456hR5drYrYrdY5wTYreYrdyerYe6y","workers":[["host04",{},29,1,"80000",0,22],["client3001",{"a":"0.27"},1,1,"80000",0,22],["host02",{"a":"0"},16,1,"80000",0,22],["host06",{"a":"0.27"},4,1,"80000",0,22],["client52",{"a":"0.27"},10,1,"80000",0,22],["host03",{"a":"0.54"},5,1,"80000",0,22],["host01",{"a":"0.54"},26,1,"80000",0,22],["host08",{"a":"0.53"},3,1,"80000",0,22],["f05",{},19,1,"80000",0,22],["client4004",{"a":"0.27"},76,1,"80000",0,22],["host05",{"a":"0.54"},36,1,"80000",0,22],["host07",{},6,1,"80000",0,22],["client5004",{},2,1,"80000",0,22],["client3002",{"a":"0.27"},7,1,"80000",0,22],["client4003",{"a":"0"},111,1,"80000",0,22],["host02",{"a":"0.54"},25,1,"80000",0,22],["client9006",{"a":"0.53"},21,1,"80000",0,22],["client6001",{"a":"0.55"},9,1,"80000",0,22],["P4003",{"a":"478.71"},1937,1,"256",0,24],["P6001",{"a":"349.75"},1936,1,"256",0,24],["p9006",{"a":"225.7"},1936,1,"128",0,24],["P5004",{"a":"369.91"},1936,1,"128",0,24],["P3002",{"a":"522.23"},1937,1,"256",0,24],["P52",{"a":"449.7"},794,1,"256",0,24],["P4004",{"a":"551.24"},1643,1,"256",0,24],["P6004",{"a":"406.18"},1936,1,"256",0,24],["P3001",{"a":"377.17"},1788,1,"256",0,24]],"algo":-1},"method":"stats.provider.workers"}

Here are in some readable view:

{"result":
    {
    "addr":"456hR5drYrYrdY5wTYreYrdyerYe6y",
    "workers":
    [
        ["host07",     {"a":"0.53"},    48,     1,  "80000",    0,  22],
        ["client52",     {},                5,      1,  "80000",    0,  22],
        ["host06",     {"a":"0.27"},    26,     1,  "80000",    0,  22],
        ["client3002",   {"a":"0"},     8,      1,  "80000",    0,  22],
        ["client4004",   {},                0,      1,  "80000",    0,  22],
        ["host08",     {"a":"0.27"},    9,      1,  "80000",    0,  22],
        ["host02",     {"a":"0.53"},    19,     1,  "80000",    0,  22],
        ["client5004",   {"a":"0.27"},  28,     1,  "80000",    0,  22],
        ["host01",     {"a":"0.27"},    16,     1,  "80000",    0,  22],
        ["client6001",   {"a":"0.53"},  45,     1,  "80000",    0,  22],
        ["client9006",   {"a":"0.53"},  26,     1,  "80000",    0,  22],
        ["host03",       {"a":"0"},     118,    1,  "80000",    0,  22],
        ["host02",       {"a":"0.27"},  78,     1,  "80000",    0,  22],
        ["f05",       {},               1,      1,  "80000",    0,  22],
        ["host05",       {"a":"0.27"},  10,     1,  "80000",    0,  22],
        ["client4003",   {"a":"0.54"},  25,     1,  "80000",    0,  22],
        ["host04",       {"a":"1.34"},  12,     1,  "80000",    0,  22],
        ["client3001",   {"a":"0.54"},  16,     1,  "80000",    0,  22]
    ],"algo":22}
,"method":"stats.provider.workers"}

I want to get names and count the number of workers in each event.
But automatically Splunk get "result.workers{}{}" field that contains all values in line:

0
1
22
80000
24
256
1986
1987
29
host02

In output I want to get table like:

Name          a          value1     value2     value3     value4     value5
---------------------------------------------------------------------------------------------------------
host07        0.53       48         1          80000      0          22
client52      0.55       51         1          80000      0          22
host06        0.27       26         1          80000      0          22
....
client3002    0          8          1          80000      0          22
0 Karma

MuS
SplunkTrust
SplunkTrust

Hi DimkoBilanko,

using the following search I was able to produce the table result you expected:

| makeresults 
| eval _raw="{\"result\":{\"addr\":\"456hR5drYrYrdY5wTYreYrdyerYe6y\",\"workers\":[[\"host04\",{},29,1,\"80000\",0,22],[\"client3001\",{\"a\":\"0.27\"},1,1,\"80000\",0,22],[\"host02\",{\"a\":\"0\"},16,1,\"80000\",0,22],[\"host06\",{\"a\":\"0.27\"},4,1,\"80000\",0,22],[\"client52\",{\"a\":\"0.27\"},10,1,\"80000\",0,22],[\"host03\",{\"a\":\"0.54\"},5,1,\"80000\",0,22],[\"host01\",{\"a\":\"0.54\"},26,1,\"80000\",0,22],[\"host08\",{\"a\":\"0.53\"},3,1,\"80000\",0,22],[\"f05\",{},19,1,\"80000\",0,22],[\"client4004\",{\"a\":\"0.27\"},76,1,\"80000\",0,22],[\"host05\",{\"a\":\"0.54\"},36,1,\"80000\",0,22],[\"host07\",{},6,1,\"80000\",0,22],[\"client5004\",{},2,1,\"80000\",0,22],[\"client3002\",{\"a\":\"0.27\"},7,1,\"80000\",0,22],[\"client4003\",{\"a\":\"0\"},111,1,\"80000\",0,22],[\"host02\",{\"a\":\"0.54\"},25,1,\"80000\",0,22],[\"client9006\",{\"a\":\"0.53\"},21,1,\"80000\",0,22],[\"client6001\",{\"a\":\"0.55\"},9,1,\"80000\",0,22],[\"P4003\",{\"a\":\"478.71\"},1937,1,\"256\",0,24],[\"P6001\",{\"a\":\"349.75\"},1936,1,\"256\",0,24],[\"p9006\",{\"a\":\"225.7\"},1936,1,\"128\",0,24],[\"P5004\",{\"a\":\"369.91\"},1936,1,\"128\",0,24],[\"P3002\",{\"a\":\"522.23\"},1937,1,\"256\",0,24],[\"P52\",{\"a\":\"449.7\"},794,1,\"256\",0,24],[\"P4004\",{\"a\":\"551.24\"},1643,1,\"256\",0,24],[\"P6004\",{\"a\":\"406.18\"},1936,1,\"256\",0,24],[\"P3001\",{\"a\":\"377.17\"},1788,1,\"256\",0,24]],\"algo\":-1},\"method\":\"stats.provider.workers\"}" 
| spath 
| table result.workers{}{} 
| rename COMMENT AS "Above code is used to create events" 
| rename result.workers{}{} AS results 
| makemv results 
| mvexpand results 
| rex field=results max_match=0 "(?<_raw>\w+\d+\s(\d+\s){4}\d+)" 
| stats values(*) AS * by _raw 
| rex "(?<host>\w+\d+)\s(?<value1>\d+)\s(?<value2>\d+)\s(?<value3>\d+)\s(?<value4>\d+)\s(?<value5>\d+)" 
| fields - _raw results

This may work for a small set of events, but can break for a larger set of events because of the mvexpand command.

I would suggest to use props.conf to create a search time field extraction using this rex

(?:\[\[|,\[)\"(?<host>\w+\d+)\",(?:\{\}|\{\"a\":\"(?<value1>[^\"]+)\"\}),(?<value2>[^,]+),(?<value3>[^,]+),\"(?<value4>[^\"]+)\",(?<value5>[^,]+),(?<value6>[^\]]+)

to create the fields and values.

Hope this helps ...

cheers, MuS

0 Karma
Got questions? Get answers!

Join the Splunk Community Slack to learn, troubleshoot, and make connections with fellow Splunk practitioners in real time!

Meet up IRL or virtually!

Join Splunk User Groups to connect and learn in-person by region or remotely by topic or industry.

Get Updates on the Splunk Community!

Federated Search for Snowflake Is Now Generally Available on Splunk Cloud Platform

Unlocking Data-In-Place Search Across Splunk and Snowflake  Enterprise data is increasingly distributed across ...

Help Us Build Better Splunk Regex Puzzles (And Win Prizes!)

If you’ve spent any time in the Splunk Community Slack, you’ve likely seen our resident Splunk Trust ...

Fuel Your Journey: What’s Waiting for You at the .conf26 Acceleration Station

Navigating the show floor at .conf26 isn't just about keynotes and technical breakout sessions; it's also ...