How to do a unique search in Splunk

Question

I'm trying to do a search in Splunk in which I'm trying to narrow it down to a unique substring.

An example of my query so far would be:

host=node-1 AND "userCache:"

Which returns something like this:

Time                Event
06/04/2021-blah     Cache miss: userCache:  tjohnson
                    host=node-1
06/04/2021-blah     Cache miss: userCache:  sbaca
                    host=node-1
06/04/2021-blah     Cache miss: userCache:  tjohnson
                    host=node-1

What I want to do, though, is to return only one unique value based on what comes after userCache:

In the above example, only two results would be returned - one for tjohnson and one for sbaca. The additional tjohnson would be stripped since there is already a tjohnson in the results.

Any suggestions?

Thanks much

Answer 1

That calls for the dedup command, which removes duplicates from the search results. First, however, we need to extract the user name into a field. We'll do that using rex.

index=foo ```Always specify an index``` host=node-1 AND "userCache:"
| rex "userCache:\s*(?<user>\w+)"
| dedup user