(RADIATOR) RewriteUsername

2000-07-01 Thread Emin TAHRALI
Title: RewriteUsername





Hi,


How can I combine this three Rewrite parameters in one expression.


RewriteUsername s/'//g
RewriteUsername tr/-A-Za-z0-9\.\@//cd
RewriteUsername s/^([^@]+).*/$1/


Best regards,
Emin TAHRALI





RE: (RADIATOR) SQL Timeout

2000-06-29 Thread Emin TAHRALI
Title: RE: (RADIATOR) SQL Timeout





Hello Hugh,


Many thanks for the given information...



The questions below can be useful to go a bit further and solve this Problem:


What are the exact conditions, when the Radioator sends TimeOut Error Message?
How does the Radioator communicate with SQL (Oracle) Server? (OK by using DBI and DBD but how?)
How does the Radiator understand, that the SQL server is down or unreachable?
For which packets does the Radiator looking for, to verify that the deletion process is OK?
Is DELETE transaction commited right after the proccess, or is there any delay?


Best Regards,


Emin TAHRALI
Siemens Business Services
mailto:[EMAIL PROTECTED]




-Original Message-
From: Hugh Irvine [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, June 28, 2000 1:40 PM
To: Emin TAHRALI; '[EMAIL PROTECTED]'
Subject: Re: (RADIATOR) SQL Timeout  SNMP Query




Hello Emin -


On Wed, 28 Jun 2000, Emin TAHRALI wrote:
 
 Hi,
 
 We are using two radiators working simultaneously, and AuthBy SQL
 authentication.
 We have some troubles with SQL Timeout Problem.
 
 The Error Logs are as below:
 
 1. Error Log of 1st Radiator
 Sun Jun 25 13:31:13 2000: ERR: do failed for 'delete from RADONLINE where
 NASIDENTIFIER='213.186.130.66' and NASPORT=090': SQL Timeout
 
 2. Error Log of 2nd Radiator
 Sun Jun 25 13:33:07 2000: ERR: do failed for 'delete from RADONLINE where
 NASIDENTIFIER='213.186.130.66' and NASPORT=090': SQL Timeout
 
 Could you please explain the conditions and the behaviour of Radiator, when
 Radiator sends these error messages.
 


This occurs because your SQL server does not respond to the SQL query shown.
When a query times out, Radiator considers that SQL server to be down and will
wait for 600 seconds by default before trying again.


Have a look at section 6.25.4 and 6.25.5 in the Radaitor 2.16.1 reference
manual.


 We guess, this problem occurs because of the SNMP query in one way or
 another, but it is not ceratin.


No. It is just because the SQL server did not respond.


 The cause of the problem can also be a deadlock in SQL or something else.
 


Yes.


 Now I need to know, when the users connection is checked by the SNMP Query.
 Below is a part of Trace 4 log. Where is the right location of the SNMP
 query
 in this sequence?
 
 And it is also not clear, why the users session is deleted before a SELECT
 query is made on the RADONLINE table.
 


What happens is this. When Radiator receives an Access-Request, it first of all
does some housekeeping and deletes any old session database record for that NAS
and Port number. This is because we might have missed a Stop record, and also
because by definition there cannot be an existing session for that NAS and Port
combination. Secondly, Radiator verifies the session database to check on
simultaneous use limits. Thirdly, only if there are already the maximum number
of simultaneous sessions for the user will Radiator then go and check with the
NAS(s) whether the sessions in the session database are still present.


 
 SSun Jun 25 17:47:27 2000: DEBUG: Handling request with Handler
 'Realm=DEFAULT'
 Sun Jun 25 17:47:27 2000: DEBUG: Rewrote user name to onurmersinlioglu
 Sun Jun 25 17:47:27 2000: DEBUG: Deleting session for onurmersinlioglu,
 213.186.131.66, 198
 Sun Jun 25 17:47:27 2000: DEBUG: do query is: delete from RADONLINE where
 NASIDENTIFIER='213.186.131.66' and NASPORT=0198
 Sun Jun 25 17:47:27 2000: DEBUG: Query is: select NASIDENTIFIER, NASPORT,
 ACCTSESSIONID from RADONLINE where USERNAME='onurmersinlioglu'
 Sun Jun 25 17:47:27 2000: DEBUG: Handling with Radius::AuthSQL
 Sun Jun 25 17:47:27 2000: DEBUG: Handling with Radius::AuthSQL
 Sun Jun 25 17:47:27 2000: DEBUG: Query is: select
 PASSWORD,CHECKATTR,REPLYATTR || SESSIONFILTER || SESSIONFILTERVALUE ||
 to_char(nvl(TOTALSESSION,2592000)) from SUBSCRIBERS,SESSIONFILTERS where
 USERNAME='onurmersinlioglu' AND
 (to_date(to_char(EXPIREDATE,'dd.mm.'),'dd.mm.')
 =to_date(to_char(SYSDATE,'dd.mm.'),'dd.mm.') OR EXPIREDATE IS NULL)
 AND subscribers.SESSIONFILTERID = sessionfilters.SESSIONFILTERID
 .


If you have defined a Simultaneous-Use CHECKATTR above, you may also check the
session database after the above query.


hth


Hugh


-- 
Radiator: the most portable, flexible and configurable RADIUS server 
anywhere. SQL, proxy, DBM, files, LDAP, NIS+, password, NT, Emerald, 
Platypus, Freeside, Interbiller, TACACS+, PAM, external, etc, etc.
Available on Unix, Linux, FreeBSD, Windows 95/98/2000, NT, MacOS X.





(RADIATOR) SQL Timeout SNMP Query

2000-06-28 Thread Emin TAHRALI
Title: SQL Timeout  SNMP Query





Hi,


We are using two radiators working simultaneously, and AuthBy SQL authentication.
We have some troubles with SQL Timeout Problem.


The Error Logs are as below:


1. Error Log of 1st Radiator
Sun Jun 25 13:31:13 2000: ERR: do failed for 'delete from RADONLINE where
NASIDENTIFIER='213.186.130.66' and NASPORT=090': SQL Timeout


2. Error Log of 2nd Radiator
Sun Jun 25 13:33:07 2000: ERR: do failed for 'delete from RADONLINE where
NASIDENTIFIER='213.186.130.66' and NASPORT=090': SQL Timeout


Could you please explain the conditions and the behaviour of Radiator, when Radiator sends these error messages.


We guess, this problem occurs because of the SNMP query in one way or another, but it is not ceratin.
The cause of the problem can also be a deadlock in SQL or something else.


Now I need to know, when the users connection is checked by the SNMP Query.
Below is a part of Trace 4 log. Where is the right location of the SNMP query
in this sequence?


And it is also not clear, why the users session is deleted before a SELECT query is made on the RADONLINE table.



Sun Jun 25 17:47:27 2000: DEBUG: Handling request with Handler 'Realm=DEFAULT'
Sun Jun 25 17:47:27 2000: DEBUG: Rewrote user name to onurmersinlioglu
Sun Jun 25 17:47:27 2000: DEBUG: Deleting session for onurmersinlioglu, 213.186.131.66, 198
Sun Jun 25 17:47:27 2000: DEBUG: do query is: delete from RADONLINE where NASIDENTIFIER='213.186.131.66' and NASPORT=0198

Sun Jun 25 17:47:27 2000: DEBUG: Query is: select NASIDENTIFIER, NASPORT, ACCTSESSIONID from RADONLINE where USERNAME='onurmersinlioglu'

Sun Jun 25 17:47:27 2000: DEBUG: Handling with Radius::AuthSQL
Sun Jun 25 17:47:27 2000: DEBUG: Handling with Radius::AuthSQL
Sun Jun 25 17:47:27 2000: DEBUG: Query is: select PASSWORD,CHECKATTR,REPLYATTR || SESSIONFILTER || SESSIONFILTERVALUE || to_char(nvl(TOTALSESSION,2592000)) from SUBSCRIBERS,SESSIONFILTERS where USERNAME='onurmersinlioglu' AND (to_date(to_char(EXPIREDATE,'dd.mm.'),'dd.mm.') =to_date(to_char(SYSDATE,'dd.mm.'),'dd.mm.') OR EXPIREDATE IS NULL) AND subscribers.SESSIONFILTERID = sessionfilters.SESSIONFILTERID

.








(RADIATOR) Acct-Terminate-Cause

2000-06-26 Thread Emin TAHRALI
Title: Acct-Terminate-Cause





Hi everybody,


In the Acct-Terminate-Cause field of Accounting Table
there are different values like 20,21,40,41,45,46,63,100,101,105,185,195,210...,
which are not defined in the dictionary file.


We use the dictioanry file with the following ID
# $Id: dictionary,v 1.26 2000/02/15 07:07:54 mikem Exp $


Now I need a document which explains these values for Acct-Terminate-Cause
I've checked the CISCO site, but there are only explanations for the values btw. 1-18.


Another question is : Is there any condition that these records come from Radiator
and not from Access Server? If so, I still need a document...



Emin TAHRALI