Splunk Search

Timechart already bucketed logs

christopherutz
Path Finder

I have a query in which I use bucket to remove some duplicates at certain intervals. I am now trying to timechart this data but I would like to use the buckets I already created rather than using the span or bins option on timechart.

When not specifying a span or bins option timechart defaults and combines the buckets split by the bucket command. Eg.

mysearch |bucket _time span=5m | dedup _time fieldA fieldB | timechart count by fieldA

If I specify a span in the timechart command it seems to work better, however, I am not convinced timechart is not still combining buckets.

mysearch |bucket _time span=5m | dedup _time fieldA fieldB | timechart span=5m count by fieldA

What are my options here?

As an example I have log entries that look like this.

timestamp fieldA=server1 fieldB=unitA 
timestamp fieldA=server1 fieldB=unitB 
timestamp fieldA=server2 fieldB=unitA 
timestamp fieldA=server2 fieldB=unitB 
timestamp fieldA=server2 fieldB=unitB 

What I am looking for is a timechart that shows the distinct count of fieldB for each value of fieldA in each bucket. So assuming all the entries above are in one bucket the value for server1 is 2 and for server2 is 2.

Tags (1)
0 Karma

sideview
SplunkTrust
SplunkTrust

Well I think if you're doing the bucketing by hand with the bucket command, you might as well go all the way and just use stats count by _time, fieldA instead of timechart count by fieldA.

The end result is the same, but when you use stats instead of timechart you'll be sure that really isnt any more bucketing going on.

UPDATE -

1) to get the "timechart showing the distinct count of field A for each value of field B", you'd just do

<everything else> | stats dc(fieldA) by _time, fieldB

2) And the flash chart doesn't care whether the rows and columns come from timechart or some manually bucketed and stats'ed search command. If the table looks right the chart will look right.

sideview
SplunkTrust
SplunkTrust

Updated my answer. Indeed the Flash Chart doesnt know what the search language was - it just graphs the rows and columns it's given, so the combination of bucket and stats dc(fieldA) by _time, fieldB will work just fine in the chart.

0 Karma

christopherutz
Path Finder

This produces the correct table but I am really interested in the chart. Can I get to the chart from stats without using timechart? I have edited the post with an example of what I am looking for.

0 Karma
Get Updates on the Splunk Community!

Stay Connected: Your Guide to May Tech Talks, Office Hours, and Webinars!

Take a look below to explore our upcoming Community Office Hours, Tech Talks, and Webinars this month. This ...

They're back! Join the SplunkTrust and MVP at .conf24

With our highly anticipated annual conference, .conf, comes the fez-wearers you can trust! The SplunkTrust, as ...

Enterprise Security Content Update (ESCU) | New Releases

Last month, the Splunk Threat Research Team had two releases of new security content via the Enterprise ...