Simon White <[EMAIL PROTECTED]> wrote: > Just maybe, there is an argument for some fallback code in the > freeradius source,
The rlm_sql module and *some* of it's drivers were updated in 0.8 to do re-connects. However, some of the drivers are not actively maintained, and weren't patched. > but somewhere in the back of my mind configurable > failover is your best bet anyway. If the downtime on your DB server is > predictable, you don't have a problem anyway. If not, get Radius to > failover to somewhere else. Instead of me re-reading configurable > failover docs, have a look yourself and come back to the list with > questions. Configurable fail-over won't help here, as the database connections will *never* come back up. Sending a HUP signal to the server may help in the short term. Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
