radsqlrelay is rude

2007-06-22 Thread Stefan Winter 
Hi,

attempting to kill a running radsql with ^C doesn't do anything, and kill'ing 
it with TERM doesn't impress it either on my system. I had to send KILL to 
get rid of it. Is this intentional?

Then, on startup I see it instantiated but only the post-auth query is printed 
on startup with -X - leaving me with the question on whether the other 
queries are actually properly read or not (see below). That's not nice as 
well. I'm on 1.1.6.

Module: Loaded sql_log
 sql_log: path = /var/log/radius/radacct/sql-relay-main
 sql_log: Post-Auth = INSERT INTO radpostauth (id, user, pass, reply, date)  
VALUES 
('', '%{User-Name}', '%{RESTENA-Service-Type}', '%{reply:Packet-Type}', '%S')
 sql_log: sql_user_name = 
 sql_log: safe-characters 
= @abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789.-_: /
Module: Instantiated sql_log (sql_relay_main)

(I'm very sure the config contains Start, Alive, Stop as well)

Greetings,

Stefan

-- 
Stefan WINTER

Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de 
la Recherche
Ingenieur Forschung  Entwicklung

6, rue Richard Coudenhove-Kalergi
L-1359 Luxembourg
E-Mail: [EMAIL PROTECTED]     Tel.:     +352 424409-1
http://www.restena.lu                Fax:      +352 422473


pgpwwyFTSNVf2.pgp
Description: PGP signature
- 
List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html

Re: radsqlrelay is rude

2007-06-22 Thread Tomas Hoger 
On 6/22/07, Stefan Winter [EMAIL PROTECTED] wrote:
 attempting to kill a running radsql with ^C doesn't do anything, and kill'ing
 it with TERM doesn't impress it either on my system. I had to send KILL to
 get rid of it. Is this intentional?

I remember hitting similar problem, when experimenting with
radsqlrelay.  IIRC I was not able to break it when it was not able to
connect do DB.  It sets $need_exit to 1 on signal, but value was not
checked when it was constantly re-trying to connect to DB.  Your
problem may be similar.

However, fixing it may cause trouble elsewhere.  From a brief look
on the script I guess you should expect duplicated records if you
break it's execution and restart it.

th.
- 
List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html

Re: radsqlrelay is rude

2007-06-22 Thread Nicolas Baradakis 
Stefan Winter wrote:

 attempting to kill a running radsql with ^C doesn't do anything, and
 kill'ing it with TERM doesn't impress it either on my system. I had to
 send KILL to get rid of it. Is this intentional?

The radsqlrelay script doesn't exit immediately. It finishes the
pending queries in sql-relay.work before that. This prevents the
next run of radsqlrelay from inserting duplicate records.

 Then, on startup I see it instantiated but only the post-auth query is
 printed on startup with -X - leaving me with the question on whether the
 other queries are actually properly read or not (see below). That's not
 nice as well. I'm on 1.1.6.
 
 Module: Loaded sql_log
  sql_log: path = /var/log/radius/radacct/sql-relay-main
  sql_log: Post-Auth = INSERT INTO radpostauth (id, user, pass, reply, date)  
 VALUES 
 ('', '%{User-Name}', '%{RESTENA-Service-Type}', '%{reply:Packet-Type}', '%S')
  sql_log: sql_user_name = 
  sql_log: safe-characters 
 = @abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789.-_: /
 Module: Instantiated sql_log (sql_relay_main)
 
 (I'm very sure the config contains Start, Alive, Stop as well)

The accounting queries are handled differently. They aren't hard-coded
in the config, that's why you can't see them in the debug output of
the parser.

-- 
Nicolas Baradakis

- 
List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html