Hello,
I have user data which is ingesting every week on Saturday in to Splunk.
I have 3000 Events on 5th Oct and 3150 Events on 12th Oct. i.e, 150 new users created in last one week.
And I have the fields called login_name and User_type.
I want to create a report showing new login_name by comparing 2 weeks of data. which is not in Splunk on 5th Oct and which is on 12th Oct.
Please do help me with the query.
Thanks in Advance.
 
					
				
		
 
		
		
		
		
		
	
			
		
		
			
					
		There probably are a few ways to do this. One is with subsearch.
index=foo earliest=-1w@w6 NOT [index=foo earliest=-2w@w6 latest=-1w@w5 | fields login_name | format]
or use set diff
set diff [index=foo earliest=-2w@w6 latest=-1w@w5] [index=foo earliest=-1w@w6]
 
					
				
		
 
		
		
		
		
		
	
			
		
		
			
					
		There probably are a few ways to do this. One is with subsearch.
index=foo earliest=-1w@w6 NOT [index=foo earliest=-2w@w6 latest=-1w@w5 | fields login_name | format]
or use set diff
set diff [index=foo earliest=-2w@w6 latest=-1w@w5] [index=foo earliest=-1w@w6]
Hello, 
 Unfortunately, I am getting all  3150 login names When I am trying with Below Query.
index=lookups sourcetype=users_roles earliest=-7d@d latest=now
 NOT 
    [search index=lookups sourcetype=users_roles earliest=-14d@d latest=-7d@d]
| stats count by login_name
3,143 events (10/7/19 12:00:00.000 AM to 10/14/19 3:21:06.000 PM)
Please do help if there is something else to sort out this issue..
Thanks,
 
					
				
		
 
		
		
		
		
		
	
			
		
		
			
					
		Have you tried the other query?
I've modified my answer. See if that makes a difference.
