Resolved issues
The following issues have been resolved in FortiADC 8.0.2 release. For inquiries about particular bugs, please contact Fortinet Customer Service & Support.
|
Bug ID |
Description |
|---|---|
| 1226241 | A system crash occurred under heavy DNS traffic on L4 UDP/TCP virtual servers using full NAT with a small SNAT pool. When the system attempted to allocate a local source port, most available ports were already in use, triggering excessive port-search loops and a resulting kernel soft lockup. |
| 1224907 | On the AWS platform, restoring a configuration file caused HA settings to be lost after reboot. The issue occurred because the HA hb-type unicast option was not parsed during the restore process. |
| 1223551 | On Hyper-V deployments, upgrading from 7.6.x to 8.0.0 or 8.0.1 caused the VM to lose its license after reboot. |
| 1221882 | In Advanced Bot Protection (ABP), JavaScript insertion failed when the HTML <head> tag included attributes (for example, <head id="xxx">). Only a plain <head></head> tag was recognized. |
| 1220203 | In Adaptive Learning, displaying a large number of URL entries could cause significant delays or GUI unresponsiveness due to the size of the data returned. Data transmission has been optimized to send large result sets in batches, improving page responsiveness when loading extensive AL statistics. |
| 1219059 | The GUI became inaccessible with a “Cannot allocate memory” error after long uptime. OSPF generated excessive error logs for virtual server IP addresses on bridge interfaces, filling the /tmp directory and preventing the GUI from loading. |
| 1218158 | Enabling compression on an L7 HTTPS virtual server together with diagnose debug for httproxy caused the process to crash due to invalid data access in the debug-printing code. |
| 1214530 | After upgrade, some VDOMs showed inactive routes and all virtual servers in those VDOMs were down. rtmd did not receive interface address and link-status updates correctly when an interface IP overlapped with a load-balance IP pool, causing the routing state to be reported incorrectly. |
| 1213980 | In a VRRP Active-Active cluster, cloning a global DNS policy with a large number of A/AAAA records caused the primary node’s Web GUI to hang and return “Bad Gateway” errors. |
| 1212271 | New CLI options are added to adjust the physical interface ring buffer size. Administrators can now configure rxring and txring values under config system interface and verify settings with the diagnose hardware get deviceinfo nic-ringsize command. |
| 1210252 | Dynamic real server pools using the Kubernetes SDN connector failed to import members when large clusters were used. Duplicate node entries and command-size limits prevented real server and pool-member configuration from being saved, causing the pool to remain empty. |
| 1208659 | HA synchronization failed when an administrator account was changed to a global user while referencing a RADIUS server configured in a non-root VDOM. The mismatch prevented the configuration from applying on the primary node and caused HA sync to stop. |
| 1205865 | The virtual-server-l2-cache-timeout setting under config router setting did not take effect in non-root VDOMs. The value was treated as a global parameter, so only the root VDOM configuration was applied. |
| 1198250 | After upgrade, all L7 TCP virtual servers failed to start and returned RST packets. The fnginx_new process was not running because it could not bind to the virtual server IP address during startup, and its reload handling did not recover from the bind failure. |
| 1049328 | SNMP queries returned empty data for VLAN interfaces due to ethtool not reporting interface statistics. |