When using the Security Services Platform (SSP) Application policy recommendation , users may experience an issue where the user interface incorrectly appends random or unrelated groups (e.g., a X group object or other default inventory items) to the intended source applications within the "Applied To" field.
SSP 5.0 , 5.1 , 5.1.1
Engineering has identified this as a UI-to-backend communication bug regarding how empty arrays are processed.
When the "Applied To" scope dialog is opened for a policy recommendation, the UI dispatches parallel queries to the backend to fetch relevant application and tier IDs.
Due to a bug in handling empty tier arrays, the backend interprets one of these parallel requests as completely unfiltered.
Consequently, the backend returns a default list of inventory items - which can be any random group from your environment depending on API pagination and naming - and the UI mistakenly merges this default list into your specifically selected applications.
A permanent fix for this array-handling bug has been developed and merged by engineering. The correction will be officially included in the upcoming SSP 5.x release.
Until you are able to upgrade to the patched 5.x release, you can safely continue using the intelligence recommendation by observing the following guidance:
Standard Publishing: You may proceed with publishing the recommended policies as usual. The extra random groups shown in the dialog are strictly a UI-level merge issue and will not be written into the actual firewall rules upon standard execution.
Precaution on Rule Modification: To avoid the potential functional impact described above, do not manually update the source or destination fields of a recommended rule before publishing. If modifications to the source/destination are necessary, publish the recommended rule first as-is, and then make the required edits to the established rule afterward to prevent the section's "Applied To" scope from erroneously overriding the rule's scope.