(RADIATOR) Duplicated Stops Accounting Records

Hugh Irvine hugh at open.com.au
Mon Oct 21 01:14:29 CDT 2002


Hello Flavio -

Duplicate radius packets are always a possibility and always a problem.

In your case, I would probably use an AcctSQLStatement instead of the 
standard AccountingTable/AcctColumnDef's, that would check to make the 
packet is not a duplicate and not already in the database. This should 
be possible by checking the username, identifier and timestamp.

Note that the timestamp in the accounting request is added by Radiator 
and is corrected for the Acct-Delay-Time automatically. You should of 
course check this.

I am fairly sure this topic has been discussed before on the mailing 
list, so have a look at the archive site and do some searching.

	http://www.open.com.au/archives/radiator

regards

Hugh


On Monday, October 21, 2002, at 12:41 AM, Ing. Flavio Rodriguez ((ISP)) 
wrote:

> Hi,
>
> I have the following problem with Radiator:
>
> Sometimes, I have more than one stop accounting record in my database 
> for user's sessions. It not happens many many times but happens. The 
> records are different between them only because they have different 
> ACCTDELAYTIME (each 60 seconds=NAS Timeout) and different TIME-STAMP 
> of course.
>
> This problem is very serious for me, because I am offering prepaid 
> cards internet access controled by time, and many stops records affect 
> the TIME-LEFT of my customers, because the TIME-LEFT is subtracted 
> many times. My billing process in general is affected, because for the 
> rest of my customers I only take data for billing process from stops 
> accounting records and if they are duplicated I have a problem ...!!!
>
> I noticed that it happens or it begins to happen when I have in my 
> logs erros like the following in my database:
>
> Sat Oct 19 09:59:57 2002: ERR: do failed for 'delete from RADONLINE 
> where NASIDENTIFIER='XXX.XXX.XXX.XXX' and NASPORT='XXXX': SQL Timeout
>
> I don't know why my database is reporting an "SQL Timeout" error. I 
> think that perhaps it can be caused for some attribute on database 
> (for example: open_cursors attribute. I think that I need to increase 
> this value (current value=)
>
> Here is my aditional information of my platform:
>
> Server: RS/6000 AIX 4.3.3 with maintenance level 10 [512 MB RAM / 9 GB 
> SCSI HARD DRIVE (whith mirrr)].
> Radiator Version: Radiator 3.3.1 whith all patches applied.
> DataBase: Oracle 8i (8.1.6 version). All the tablespaces have enought 
> disk free.
>
> Also here is a part of my logs when it happens:
>
> Sat Oct 19 09:15:22 2002: ERR: do failed for 'delete from RADONLINE 
> where NASIDENTIFIER='XXX.XXX.XXX.XXX' and NASPORT='XXXX': SQL Timeout
> Sat Oct 19 09:16:40 2002: INFO: Duplicate request id 158 received from 
> XXX.XXX.XXX.XXX(YYYY): ignored
> Sat Oct 19 09:16:40 2002: INFO: Duplicate request id 158 received from 
> XXX.XXX.XXX.XXX(YYYY): ignored
> Sat Oct 19 09:16:40 2002: INFO: Duplicate request id 158 received from 
> XXX.XXX.XXX.XXX(YYYY): ignored
>
> Any help will be appreciated.
>
> Regards.
>
> Eng. Flavio Rodriguez
> ISP - ETAPA
>

NB: I am travelling this week, so there may be delays in our 
correspondence.

-- 
Radiator: the most portable, flexible and configurable RADIUS server
anywhere. Available on *NIX, *BSD, Windows 95/98/2000, NT, MacOS X.
-
Nets: internetwork inventory and management - graphical, extensible,
flexible with hardware, software, platform and database independence.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 3327 bytes
Desc: not available
URL: <http://www.open.com.au/pipermail/radiator/attachments/20021021/8fb77051/attachment.bin>


More information about the radiator mailing list