Hi Alan, Thank you very much for responding so quickly.
I'm using the EXACT same radtest parameters (same user/pass etc) with every attempt. It's only every third Access-Request that fails. > > rlm_sql: Failed to create the pair: Unknown value for attribute > > ChilliSpot-Max-Output-Octets > > rlm_sql (sql): Error getting data from database > > Why not fix that, then? It looks like you've added an attribute > with no value into SQL. Why did you do that? I'm not sure how to fix that. It only happens every third time. I didn't add any attributes with no value into SQL. Note: I have Simultaneous-Use set to 1 for this user (let's call him User X). (It's essential that it be only one). For a user WITHOUT a Simultaneous-Use record (User Y) the behaviour is a bit different: At first it LOOKS as though User Y never has this problem... except randomly... At further inspection though: If I switch from radtesting User X to radtesting User Y, on the THIRD attempt, I will get a Reject on User Y. After that all radtests on User Y will return Accept. What am I doing wrong? ------------------------------------------------------------- On 6/14/06, Stelio Gouveia <[EMAIL PROTECTED]> wrote: > Interesting response > > > > ---------- Forwarded message ---------- > From: "Alan DeKok" <[EMAIL PROTECTED]> > To: FreeRadius users mailing list <freeradius-users@lists.freeradius.org> > Date: Wed, 14 Jun 2006 11:23:03 -0400 > Subject: Re: The Accept Reject 123 Waltz > Stelio Gouveia <[EMAIL PROTECTED]> wrote: > > EVERY third radtest packet I sent gets rejected! 123! 123! 123! Every > > now and then there's a bit of randomness, e.g. two rejects in a row, > > but I can sit in front of the computer all day with the up arrow and > > dance to the rythm! > > I've never seen that before... > > > rlm_sql: Failed to create the pair: Unknown value for attribute > > ChilliSpot-Max-Output-Octets > > rlm_sql (sql): Error getting data from database > > Why not fix that, then? It looks like you've added an attribute > with no value into SQL. Why did you do that? > > It looks like in one case, SQL is returning a password, and in > another, it's not. > > Alan DeKok. > > - > List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html > > > > __________ NOD32 1.1597 (20060613) Information __________ > > This message was checked by NOD32 antivirus system. > http://www.eset.com > > > > > > - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html