I'm running FreeRadius 2.1.12 on RedHat 6 as an Eduroam proxy server and having problems. My cisco wireless controllers are constantly failing back and forth between the primary and secondary systems. My logs show these errors when it happens: Dec 4 19:12:18 server radiusd[32588]: Internal sanity check failed for child state Dec 4 19:12:18 server radiusd[32588]: Reply from home server x.x.x.x port 1812 - ID: 210 arrived too late for request 76479. Try increasing 'retry_delay' or 'max_request_time' Dec 4 19:13:18 server radiusd[32588]: Discarding duplicate request from client ct5508 port 32770 - ID: 148 due to unfinished request 76495 Dec 4 19:13:26 server radiusd[32588]: Discarding duplicate request from client ct5508 port 32770 - ID: 148 due to unfinished request 76495 Dec 4 19:13:34 server radiusd[32588]: Discarding duplicate request from client ct5508 port 32770 - ID: 148 due to unfinished request 76495 Dec 4 19:13:41 server radiusd[32588]: Internal sanity check failed for child state Dec 4 19:13:41 server radiusd[32588]: Reply from home server x.x.x.x port 1812 - ID: 102 arrived too late for request 76495. Try increasing 'retry_delay' or 'max_request_time' Dec 4 19:13:42 server radiusd[32588]: Discarding duplicate request from client ct5508 port 32770 - ID: 148 due to unfinished request 76495 I'm look for some advice as to what timeouts to adjust that might help with the errors. I didn't want to start changing retry_delay or max_request_time without some advice first. Any suggestions? I have no control over the Home Servers or how long a request takes to complete and I don't know what's causing the sanity check failure!? -Mike