Hi Everyone I've been trying to setup MySQL fail over with freeradius I've followed http://wiki.freeradius.org/SQL_HOWTO#Additional_Snippets But I get weird results. If I stop the first sql server service, freeradius starts to use the second sql server, as expected. But if I stop the entire first server (ie poweroff) freeradius still continues to try and use sql1, hanging... I have tryed this with real servers and with virtual servers, both react the same, if you take down the first sql service it will start to use the second, but if you power off or suspecd the first sql server, radius hangs trying to get to the server. I have tryed freeradius 2.1.0., 2.1.1 and 2.1.3, on redhat enterprise 4 compiled from source. Restarting freeradius it will notice the first sql server is down and use the second sql server. Can anyone confirm this behaviour? Any ideas how to get it working correctly? Secondly, I have been looking into sqlippool and MySQL fail over. We would really like to be able to use a read/write master and read only slave, but it looks to me that the sqlippool needs to be writeable to mark the IP address as used and avoid duplicate IP allocation. Can someone confirm this? Thanks Peter