Re: sql_ippool module with two databases

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Re: sql_ippool module with two databases

Ruslan A Dautkhanov
Hello

Stylianos Stylianou wrote:

Hi,

 

I have compiled freeradius-1.0.5 in Redhat Enterprise Linux 4 with your module and it works without any problem with mysql 4.X database.

 

I have only one problem to solve. I am trying to add two sql-instance-name in file sqlippool.sqls so that there is a redundancy in the module.

In the radius.conf file I manage to do this with the following:

 

group {

        sql_1 {

              fail  = 1

              notfound = return

              noop  = 2

              ok  = return

              updated = 3

              reject = return

              userlock = 4

              invalid = 5

              handled = 6

             }

        sql_2 {

              fail  = 1

              notfound = return

              noop  = 2

              ok  = return

              updated = 3

              reject = return

              userlock = 4

              invalid = 5

              handled = 6

            }

}

 

So, if sql_1 server goes down, the authentication and accounting is going to be ok since radius is going to use sql_2 server, but the allocation of ips not.

 

Is there a way to “tell” the module to use two sql servers, the one redundant to the other?

 

Current revision of sqlippool can't make this functionality, as I know.   To be sure ask for it at [hidden email] .
If it's really not possible, then only one solution possible - search for utilities for your RDBMS, which can make redundancy.
I know it's possible in PostgreSQL, but I havn't any experience with MySQL.

--
best regards,
Ruslan A Dautkhanov    [hidden email]


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

smime.p7s (4K) Download Attachment