Sample data set
user, pc, logon, logoff, durationOfLogon
User11, HNA1E8I, 01-06-15 13:49:09, 01-06-15 13:49:11, 0:00:02
User138, HN2OTS12, 05-06-15 13:02:33, 05-06-15 13:02:36, 0:00:03
User160, H3NB3RA, 05-06-15 10:29:14, 05-06-15 10:29:16, 0:00:02
User174, HN4WF101, 03-06-15 11:55:43, 03-06-15 11:55:46, 0:00:03
User237, 3HND5, 01-06-15 11:00:21, 01-06-15 11:00:29, 0:00:08
User237, 3HND5I22, 01-06-15 11:00:22, 01-06-15 11:00:23, 0:00:01
User36, 4HNAE3I, 03-06-15 12:47:46, 03-06-15 12:47:48, 0:00:02
User379, 4HND5A, 03-06-15 11:19:40, 03-06-15 11:19:40, 0:00:00
User386, 4HNE4A, 07-06-15 16:33:51, 07-06-15 16:33:53, 0:00:02
User62, 6HNAE3I, 03-06-15 18:32:31, 03-06-15 18:32:34, 0:00:03
User640, 34HNF4, 01-06-15 10:04:19, 01-06-15 10:04:21, 0:00:02
User640, 5NF4, 04-06-15 16:04:18, 04-06-15 16:04:20, 0:00:02
User747, NOCC01, 01-06-15 12:37:02, 01-06-15 12:37:03, 0:00:01
User747, 23NOCC03, 02-06-15 8:42:20, 02-06-15 8:42:26, 0:00:06
User747, ATPH02, 02-06-15 8:42:23, 02-06-15 8:42:25, 0:00:02
User790, 4HN2Y12, 03-06-15 15:29:33, 03-06-15 15:29:35, 0:00:02
User790, 3HNPHY02, 03-06-15 15:29:33, 03-06-15 15:29:40, 0:00:07
How can I create an search to find out the users with concurrent access on different PCs within the logon and logoff period?
Thanks
Kelvin
You can do this by splitting each event into a logon event and a logoff event, assuming logon time is _time:
sourcetype=logons | eval count=1 | appendpipe [| eval _time = strptime(logoff, "%d-%m-%y %H:%M:%S") | eval count=-1] | sort + _time | streamstats sum(count) as concurrency by user global=f | where concurrency > 1
Hi dturnbull,
Thanks for your reply.
However, the query seems not working on my situation. The result is not match when compare with the raw data. Here is the search query - "index=sample1 UserId!="User" | eval count=1 | appendpipe [ | eval _time = strptime(LogoffTime, "%d-%m-%Y %H:%M:%S") | eval count=-1 ] | sort + _time | streamstats sum(count) as concurrency by UserId global=f | where concurrency > 1 | table Site UserId WorkStationId LoginTime LogoffTime | sort 0 + Site"
The way i am using now is "JOIN", but this is very slow and i know might have some smart way to do it.
Thanks for your help
Kelvin