freeradius2 Load-Balance Issue
Hi, I am using FreeRADIUS Version 2.1.7, this is the RPM for RHEL made available from (http://people.redhat.com/jdennis/freeradius-rhel-centos) which was linked by the RedHat FAQ on the FreeRadius Wiki (actually I downloaded the .srpm and change the .spec file to include --with-udpfromto in the configure line). Anyway I seem to have an odd issue. I have two home servers configured with the below proxy.conf, the issue is when both server1 and server2 are up load balancing works fine when I take down server 1 it goes into zombie then it is marked dead which is all great (I posted those messages below as well), then when I bring server 1 back online FreeRadius' status_check sees the server up ( I posted that message below as well) however no more Radius requests go to server1 (not unless I take server2 down then I can see server1 starts to receive the requests) but it seems like either there is a bug or I have something mis configured because I would assume once the status_check notices server1 is back online it should start to receive radius requests again.. home_server server1 { type = auth+acct ipaddr = XXX.XXX.XXXX.XXXX port = 1812,1813 secret = secret require_message_authenticator = no response_window = 20 zombie_period = 10 # revive_interval = 120 status_check = request username = username@host.com password = secret check_interval = 5 num_answers_to_alive = 3 } home_server server2 { type = auth+acct ipaddr = XXX.XXX.XXX.XXX port = 1812,1813 secret = secret require_message_authenticator = no response_window = 20 zombie_period = 10 # revive_interval = 120 status_check = request username = username@host.com password = secret check_interval = 5 num_answers_to_alive = 3 } home_server_pool balance { type = load-balance home_server = pcrf01 home_server = pcrf02 } realm DEFAULT { pool = balance nostrip } ------------------ end proxy.conf -------------------- ------ DEAD SERVER -------------------- PROXY: Marking home server XXX.XXX.XXX.XXX port 1812 as zombie (it looks like it is dead). Sending Access-Request of id 222 to XXX.XXX.XXX.XXX port 1812 User-Name := "username@host.com" User-Password := "secret" Service-Type := Authenticate-Only Message-Authenticator := 0x00000000000000000000000000000000 NAS-Identifier := "Status Check. Are you alive?" Waking up in 3.9 seconds. Waking up in 4.3 seconds. ---------- end DEAD SERVER ------------ ------ Server Alive ----------------- rad_recv: Access-Accept packet from host XXX.XXX.XXX.XXX port 1812, id=185, length=20 Waking up in 10.0 seconds. --------- end Server Alive -------------
Brian Carpio wrote:
I am using FreeRADIUS Version 2.1.7, this is the RPM for RHEL made available from (http://people.redhat.com/jdennis/freeradius-rhel-centos) which was linked by the RedHat FAQ on the FreeRadius Wiki (actually I downloaded the .srpm and change the .spec file to include --with-udpfromto in the configure line).
Try 2.1.8. It fixes some odd issues when proxying to home servers that are down. Alan DeKok.
Yes thank you very much upgrading to 2.1.8 resolved the issue. Brian On Wed, Dec 30, 2009 at 10:12 AM, Alan DeKok <aland@deployingradius.com>wrote:
Brian Carpio wrote:
I am using FreeRADIUS Version 2.1.7, this is the RPM for RHEL made available from (http://people.redhat.com/jdennis/freeradius-rhel-centos) which was linked by the RedHat FAQ on the FreeRadius Wiki (actually I downloaded the .srpm and change the .spec file to include --with-udpfromto in the configure line).
Try 2.1.8. It fixes some odd issues when proxying to home servers that are down.
Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
participants (2)
-
Alan DeKok -
Brian Carpio