Hi there,
I wonder if Riak is a good way to do what I want. And I would ask her if my
design is possible.
The goal is to be able to get the access sum on a given address againts a
given period in the logs of my web server.
Each days I get the log file of the day before ( 4249421 lines for 1 server
).
compute it to create a sum of line number for each URL and create a file
with json insertable lines (about 200000 each days for a platform) looks
like
{ "Count": 15, "Date":"20100813", "URL":"/home", "Method":"GET" }
and inject them in my bucket 'PlatformLog' with random key
Does it sounds good/weird ?
my example of map/reduce (REST):
{
"inputs": "PlatformLog",
"query": [
{
"map": {
"language": "javascript",
"source": "function(v,k,a) { var
data=JSON.parse(Riak.mapValuesJson(v)[0]); return (data.Time > a[1] &&
data.Time > a[2] && data.URL == a[0]) ? [data.Count] : [] ; }",
"keep": false,
"arg": [ "/home", "20100801", "20100901" ]
}
},
{
"reduce": {
"language": "javascript",
"name": "Riak.reduceSum"
}
}
],
}
I suppose a great enhancment would be keys based on date to make some more
filter in the input fields.
Can I insert a whole object table of 200000 items as
http://server/PlatformLog/20100813 ?
How can I manage selection of keys ? if one does not exists in my range of
date for example.
Thank you for your advises.
Best regards,
--
Damien
_______________________________________________
riak-users mailing list
[email protected]
http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
