Splunk Search

Need help with json spath search

madhav_dholakia
Contributor

Hello, 

I have seen a few of the spath topics around, but wasn't able to understand enough to make it work for my data. 

I would like to create a line chart using pointlist values - it contains timestamp in epoch and CPU%

Search I tried but not working as expected to extract this data:

index="splunk_test" source="test.json"   
| spath output=pointlist path=series{}.pointlist{}{}
| mvexpand pointlist
| table pointlist

madhav_dholakia_0-1704878418188.png

Please see below sample json.

{"status": "ok", "res_type": "time_series", "resp_version": 1, "query": "system.cpu.idle{*}", "from_date": 1698796800000, "to_date": 1701388799000, "series": [{"unit": [{"family": "percentage", "id": 17, "name": "percent", "short_name": "%", "plural": "percent", "scale_factor": 1.0}, null], "query_index": 0, "aggr": null, "metric": "system.cpu.idle", "tag_set": [], "expression": "system.cpu.idle{*}", "scope": "*", "interval": 14400, "length": 180, "start": 1698796800000, "end": 1701388799000, "pointlist": [[1698796800000.0, 67.48220718526889], [1698811200000.0, 67.15981521730248], [1698825600000.0, 67.07217666403122], [1698840000000.0, 64.72434584884627], [1698854400000.0, 64.0411289094932], [1698868800000.0, 64.17585938553243], [1698883200000.0, 64.044969119166], [1698897600000.0, 63.448143595246194], [1698912000000.0, 63.80226399404451], [1698926400000.0, 63.93216493520908], [1698940800000.0, 63.983679174088145], [1701331200000.0, 63.3783379315815], [1701345600000.0, 63.45321248782884], [1701360000000.0, 63.452383398041064], [1701374400000.0, 63.46314971048991]], "display_name": "system.cpu.idle", "attributes": {}}], "values": [], "times": [], "message": "", "group_by": []}

can you please help how I can achieve this?

Thank you.

Regards,

Madhav

Labels (3)
0 Karma
1 Solution

ITWhisperer
SplunkTrust
SplunkTrust

It appears that two dimensional arrays are not easily handled (unless someone else knows differently), so you could try something like this:

| spath output=pointlist path=series{}.pointlist{}{}
| mvexpand pointlist
| table pointlist
| streamstats count as row
| streamstats count(eval(row % 2==1)) as row
| stats list(pointlist) as pointlist by row
| sort 0 row
| eval pointX = mvindex(pointlist,0)
| eval pointY = mvindex(pointlist,1)

View solution in original post

0 Karma

ITWhisperer
SplunkTrust
SplunkTrust

It appears that two dimensional arrays are not easily handled (unless someone else knows differently), so you could try something like this:

| spath output=pointlist path=series{}.pointlist{}{}
| mvexpand pointlist
| table pointlist
| streamstats count as row
| streamstats count(eval(row % 2==1)) as row
| stats list(pointlist) as pointlist by row
| sort 0 row
| eval pointX = mvindex(pointlist,0)
| eval pointY = mvindex(pointlist,1)
0 Karma

madhav_dholakia
Contributor

Hi @ITWhisperer - thanks a lot, this worked like a charm.

0 Karma
Get Updates on the Splunk Community!

Fall Into Learning with New Splunk Education Courses

Every month, Splunk Education releases new courses to help you branch out, strengthen your data science roots, ...

Super Optimize your Splunk Stats Searches: Unlocking the Power of tstats, TERM, and ...

By Martin Hettervik, Senior Consultant and Team Leader at Accelerate at Iver, Splunk MVPThe stats command is ...

How Splunk Observability Cloud Prevented a Major Payment Crisis in Minutes

Your bank's payment processing system is humming along during a busy afternoon, handling millions in hourly ...