Splunk Search

Don't get eval based macros

andersmholmgren
Explorer

I just can't seem to understand how the eval based macros are supposed to work

I wrote a very simple macro

[TEST]
definition = "TEST"
iseval = 1

then a query to test the output

index=_audit | head 1 | eval test1=`TEST`  | eval test2=tostring(`TEST`) | table test*

The output is one column 'test2' with a value of Null

Why is that? Shouldn't the value be "TEST" for both columns? If not why not?

Tags (3)
0 Karma
1 Solution

genthaler
Engager

Try this:

[TEST]
definition = "\"TEST\""
iseval = 1

View solution in original post

gkanapathy
Splunk Employee
Splunk Employee

An eval-based macro returns a string, which is substituted into the query. Your macro returns the string TEST, without quotes, so you are getting:

... | eval test1=TEST  | eval test2=tostring(TEST) | ...

In this case, TEST is used as the name of a non-existent variable. You can get what you intended either with @genthaler's answer, or by putting the quotes in the query:

... | eval test1="`TEST`"  | eval test2=tostring("`TEST`") | ...

genthaler
Engager

Hi @gkanapathy,
I just tried it, unfortunately quoted macro invocations don't get invoked.
So instead of "TEST", you end up with the literal string "`TEST`".

0 Karma

genthaler
Engager

Try this:

[TEST]
definition = "\"TEST\""
iseval = 1

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!

Make Your Machine Data AI-Ready in Splunk

Every team is getting ready for AI agents.   Thursday, November 12, 2026  |  11AM  PST / 2PM EST   ...

Splunk Asynchronous Forwarding Explained

Splunk asynchronous forwarding is often misunderstood as simply setting autoLBVolume. That is not quite right. ...

How to Monitor AI Agents with Splunk Agent Observability

Let’s say I’m running a travel planning AI app in production. A user asks for three concise hotel options in ...