$ cd
$ lw-compute
Usage: /usr/local/bin/lw-compute [options]
-g group Specify an aggregation group (e.g, hour(_ts_utc), or response.code)
-c condition Specify a condition to match. May be used multiple times
-s starttime Specify the start time for the query
-e endtime Specify the end time for the query
-v Show debug information
-h, --help Display this screen
Based on the web_log log table, which is created automatically when you use logworm:
$ lw-compute web_log count
==> Get the total count of logentries in the web_log log
count = 30420.0
$ lw-compute web_log sum profiling
==> Get the total amount of time spent processing requests
==> (profiling is the field that records the number of ms taken by the request)
sum(profiling) = 746.503927999989
$ lw-compute web_log count -g response_status
==> Count the number of requests by response status (eg, 200, 201, 300, 302, 304, 404, 500)
[response_status:200] ==> count = 71.0
[response_status:304] ==> count = 17.0
[response_status:404] ==> count = 2.0
[response_status:500] ==> count = 1.0
$ lw-compute web_log count -g response_status -s 2010-03-01
==> Count the number of requests by response status including only requests since March 1st
$ lw-compute web_log count -g response_status -s 2010-02-01 -e 2010-02-28
==> Count the number of requests by response status including only requests in February
$ lw-compute web_log count -g response_status -s 2010-02-01 -e 2010-02-28 -c '"request_path":"/foo"'
==> Count the number of requests of /foo by response status including only requests in February