Do we need to use a subsearch to divide the results of one search by the results of another search to find a percentage?

[splunker9999]

Hi,

We are looking for timeout percentage from the total events.

For Ex:
1. Query1: index=datapower Time=*|stats count

2. Query2: index=datapower Time>10000|stats count

Now we are looking for results as Query1/Query2. Do we need to do this with subsearch? Can anyone help with this?

Thanks

[jrodman]

The strict answer to your question is yes, you need two searches for the general case, whether subsearches, or other methods like reusing the output of one search as input in another.

However the usual answer is to craft a search that doesn't need to handle the general case, and thus this is a comment since +somesoni2 has addressed the inquiry.

[somesoni2 · Accepted solution]

Try something like this

index=datapower Time=* | stats count as Query1 count(eval(Time>10000)) as Query2 | eval Result=Query1/Query2

OR

index=datapower Time=* | eval Query2=if(Time>10000,1,0) | stats count as Query1 csum(Query2) as Query2 | eval Result=Query1/Query2