multiple searches for stats

[kishan2356]

I have two searches where I need to run an stats count on to do some calculations. First search  is

index=xxx wf_id=xxx wf_env=xxx xxx | stats count

Second search is 

index=xxx wf_id=xxx wf_env=xxx    sourcetype=xxx usecase=xxx  | stats count by request_id

First search uses a simple stats to get its count, but the second search uses stats count by request_id so I am having trouble getting the counts for both. Ideally I would like to get the counts for both searches and divide them. I've used appendcols but it returns empty fields for both searches. Any guidance on how to get counts for these searches would be helpful! 

Working example:

_time	Search 1 counts	Search 2 counts	Search 1/ Search 2
00:30	50	25	2
00:35	100	25	4

[kishan2356]

This returns an empty field 

[gcusello]

Hi @kishan2356,

in my test I have two rows: one with empty field and one with the valued fiel, you could filter for the fields with the value:

index=xxx wf_id=xxx wf_env=xxx sourcetype=xxx usecase=xxx
| stats count by request_id
| appendpipe [ search index=xxx wf_id=xxx wf_env=xxx xxx | stats count | rename count AS total ]
| search total=*
| eval perc=count/total

Ciao.

Giuseppe

[gcusello]

Hi @kishan2356,

did you tried appendpipe?

something like this:

index=xxx wf_id=xxx wf_env=xxx sourcetype=xxx usecase=xxx
| stats count by request_id
| appendpipe [ search index=xxx wf_id=xxx wf_env=xxx xxx | stats count  rename count AS total ]
| eval perc=count/total

Ciao.

Giuseppe

[kishan2356]

Thank you for the reply Giuseppe,

Hi am still getting black fields when using appendpipe.

[gcusello]

Hi @kishan2356,

please try this:

index=xxx wf_id=xxx wf_env=xxx sourcetype=xxx usecase=xxx
| bin _time span=5m
| stats count by _bin request_id
| append [ search 
   index=xxx wf_id=xxx wf_env=xxx xxx 
   | bin _time span=5m 
   | stats count BY _time 
   | rename count AS total 
   ]
| stats sum(count) AS count values(total) AS total BY _time
| eval perc=count/total

Ciao.

Giuseppe