Input: Apache Kafka with Ngnix log(milliseconds level) from multiple servers, the logs is generally ordered by time, but not exactly.
Proceed: What tools or tech. should I choose ???
Output: Text with format: URL time(minute level) count
e.g.
Input: 2015-06-24 12:23:34.001 /rest/login?user=xxx HTTP/1.0 ...
Output: /rest/login "2015-06-24 12:23" 2
Thanks very much for your reply
BTW. parsing the log is not the problem, but how to do the aggregation and count.