Dave; fair question and I get a lot of variations on this i.e. why do I see OCONUS IP data in my logs etc. There are numerous reasons for this that differ based on the product or service in question. At a high level there are 3 categories of issues. First is telemetry. We manage a global service fabric and there is telemetry infrastructure worldwide to support that. Often there are system level issues being logged into the most responsive infrastructure; other times one region is checking the health, performance or for other updates to validate across the global fabric. In each of these scenarios it is important to note that in our commitments to government services we commit we will not send or store your content outside the accreditation boundary. These interactions are limited to what we classify as system data. We've invested in, and implemented, significant automation to ensure data is appropriately 'scrubbed' before it's sent to or logged into any of our underlying service repositories. Second; there are a category of issues that fit into more complex 'by design' scenarios where analysis shows that users were engaging with services, content, etc. across regions which incurred logging of OCONUS addresses etc. Lastly there have been defects at times that erroneously log data that creates false positives. In all examples our commitment is to ensure we maintain the customer data within the accreditation boundary. We cannot prevent customers from cross regional interaction; and many government customers have justifiable OCONUS missions requiring such. Ultimately this becomes a balancing act where a black/whitelist; geofencing, etc. can have unintended performance requirements. Over time I have found data layer protections much more effective than network layer due to the increasingly complex dependencies on global infrastructure as well as increasing global composition of workforces and missions. By implementing a fuller range of ZT (zero trust) capabilities we've seen much more effective protections evolving at pace with more dynamic traffic needs. I do sympathize with the challenges having been there myself in previous roles. You should be able to engage your support team to help you find answers specific to the scenarios you encounter; any recommendations we may have; where geofencing or other constraints are known to create possible performance problems etc.