Different ACL behavior in different nodesDescriptionWhen attempting to read/write/delete/create records from different instance nodes, the ACL behavior varies from one node to the next.CauseThis can be caused by caching after an instance upgrade. The ACL is not being recognized when the ACL execution plan is being evaluated, and so the execution plan is falling back on the global (*) ACLs.ResolutionTo verify the ACL is not included in the execution plan, you can navigate to the sys_security_acl record that should allow the user to read/write/create/delete the record, then click the 'Show ACL Execution Plan' related link. If the modal that appears is not showing the current ACL as part of the execution plan: Deactivate the ACLSave the recordReactivate the ACLSave the recordVerify the ACL shows in the 'Verify Security Rules' popup window that is shown.