Diff the counts in correlated queries

[goalkeeper]

I am very new to Splunk.
I have two log files, the first one,  let's call it accessLog, contains the access log for the http requests

A splunk query could give me the count for each request.

url                                        count
http://host1/query1      10
http://host1/query2      20

The second log file, let's call it errorLog, contains only error for the request, the line contains a keyword of the url.

A Splunk query could give me the result:

keyword           errorCount
query1              2
query2              8

I want to calculate the success ratio for each request:

URL                                                success ratio
http://host1/query1              80%
http://host2/query2              60%

It could be described as the following sql

Select url,  count(url),   (count(url) - (select (count (keyword) from ErrorLog where url is like '%keyword%'))/count(url) as successRatio
From accessLog
group by url

Could this be done in a Splunk query?

Thanks in advance.

[to4kawa · Accepted solution]

| makeresults
| rename COMMENT as "First search"
| eval _raw="url    count
http://host1/query1      10
http://host1/query2      20" 
| multikv forceheader=1 
| table url count 
| append 
    [| makeresults
    | rename COMMENT as "Second search" 
    | eval _raw="keyword           errorCount
query1              2
query2              8" 
    | multikv forceheader=1 
    | table keyword errorCount] 
| eval tmp=mvindex(split(url,"/"),-1) 
| eval key=coalesce(keyword,tmp) 
| selfjoin key 
| eval "success ratio"=(1-errorCount/count)*100 
| table url "success ratio"

But

index=yours (sourcetype=AccessLog OR (sourcetype=ErrorLog "keyword"))
and aggregate with stats command
This is Splunk way.

[goalkeeper]

Works like a charm. thanks again.