On my 'Threat Activity Dashboard', I see a panel labeled 'Most Active Threat Collections', but the numbers don't seem to change. I would think this panel should include all of the threat collections and then graphically show me which are the most active. I'm new to Splunk so I'm trying my best to figure it out and could use a little help. Here is what is currently in that panel (which was set up before my time).
| tstats `summariesonly` values(source) as search,values(Threat_Activity.threat_collection_key) as threat_collection_key,count from datamodel=Threat_Intelligence.Threat_Activity where * NOT NOT () by _time,Threat_Activity.threat_collection span=1h | `drop_dm_object_name("Threat_Activity")` | stats values(search) as search,sparkline(sum(count),1h) as sparkline,dc(threat_collection_key) as dc(artifacts),sum(count) as count by threat_collection | rex max_match=10 field=search "Threat\s*-\s*(?<friendly_search>.*)(?:\s*-\s*Threat\s+Gen)" | rename friendly_search as search | sort - count
Thanks in advanced!
The values that are in the search here tie back to matches between threats and log events. The threat artifact dashboard is where all the threat collections reside, but threat activity shows matches with your data. The search you saw is the default threat collection search. dc(artifacts) is the count of threat artifacts being correlated to events. The count ties to the distinct logs returned.
Ahh okay, so that panel is the data we have in our Splunk (our logs) correlated or matched with threat intelligence artifacts. In the panel I only see one 'threat_collection' category return and that is 'ip_intel', should I be seeing more than that? I understand this is a pretty broad question because you can't see my logs or dashboard, I guess I'm just wondering if the search it is performing is sufficient.
Thanks for your help btw.
It would really depend on the data sets you are collecting and the threat feeds that you have downloaded. There are a set of background saved searches that run to get the data pulled together so if they are not firing then it should be alright. You could test this by taking some known data and known indicators, perhaps a test system and put them into local_ip or other local lists and see what comes out the other side to validate further.