On Mon, May 17, 2004 at 09:03:47PM +0200, Michael Markstaller wrote: > running FreeRADIUS Version 1.0.0-pre0 (Debian-package from cvs20040421) > with accounting to mysql 3.23.49. > When I understood things right, the "accounting_update_query_alt" should > come into action when "accounting_update_query" fails because no > start-record was ever inserted, right ?
> However, if I'm right ;) when the start-record gets recorded, accounting > including updates works fine, but when the according row is missing in > table radacct (due to the start-record got lost or accounting started > after the session was created - ok, this is really uncommon) then > "accounting_update_query_alt" doens't kick in. > From what I understood the problem lies in rlm_sql not reporting an > error somehow when the update fails; complete debug-output from one > update and the stop-record below (both with changed IP's. > The stop-record finally creates a row in radacct, so > "accounting_stop_query_alt" works fine. > As soon as the according start-entry in table radacct exists also the > update work fine, but I thought of this as a way to recover missing > accounting packets to at least have the updates creating records when > the start goes wrong. Yes, it was a mistake on my part.... An update that affects no rows is not an error. I took the code from the start handling, instead of from the stop handling. It should be fixed now, in rlm_sql.c 1.130. -- Paul "TBBle" Hampson, on an alternate email client. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
