Forum Discussion
Trying to understand "Anomalous sign-in location by user account and authenticating application"
Microsoft
You can use the Query
SigninLogs
| where TimeGenerated > ago(14d)
| where UserPrincipalName == "...."
| extend locationString = strcat(tostring(LocationDetails["countryOrRegion"]), "/", tostring(LocationDetails["state"]), "/", tostring(LocationDetails["city"]), ";")
summarize count() by locationString
Thanks for the answer.
This gave a good view of the sign-ins.
But are these successfull sign-ins or does it also count the failed sign-ins?
- Ofer_Shezaf
It is all sign ins because I kept to the data as analyzed by the rule itself. If you want to investigqate Signins with more flexiblity, you can use the Signins workbook which is very useful. You can learn more about this workbook (as well as the sign-in mapping workbook) in the "Day in a SOC analyst life" webinar (see module 12 of https://aka.ms/sentinelninjatraining)
Okay, i will look into that.
Im just trying to understand why the incident is actually being created and what the response should be to it.Thanks for the help
- shainw
mircasa - Thanks for the feedback. I am looking at the detection and we will likely have some updates in the next week available on the Azure Sentinel GitHub. The involved App should already be coming thru in the AppDisplayName, but agreed we should bring thru the Location information, the ResultType for the sign in (meaning success or fail error code), along with IPAddresses related to the UserPrincipalName that is making the Signin attempt. The goal of this detection is to indicate a UserPrincipalName for a given AppDisplayName is anomalous based on the location the IP is associated with, all relative to the last day, 7 days and 14 days. If an alert fires for this, then using the workbook that Ofer points out would be a next step to understand context for the user and Signins. We can also look at improving the description to help with this. I will post back once the new version is available.
