Load Balancing Problem



  • Load balancing and Fail over was working and now it is not.

    I was using an earlier version of the 2.0 Beta 4 build from sometime in September. But in an attempt to fix the problem I updated the system on October 5th so now I am now using 2.0-BETA4  (i386) built on Tue Oct 5 11:58:55 EDT 2010. After installing the update I setup the routing groups and added it to the firewall rules like I did before, but this did not fix the problem.

    The 2 changes that I made when the problem began are.

    1.   I changed my 2 DSLs that are used for the load balancing from Dynamic to a PPoE configuration, so PFSense would have a public IP address rather than a private address which was necessary for what I am doing.

    2.   Another difference from when it worked and now is that WAN 2 is setup as default. Originally to get load balancing to work I had to remove the “default gateway” designation from any of the WANs, under routing. I am not sure if this is something that should be done or not but when I did it originally it worked so long as I added the proper gateways or gateway group to the firewall rules.

    When I tried doing this again after changing my 2 WANs to PPoE instead of Dynamic it corrupted the gateways similar to what Casinmirad described on bullet point number 2 of his first post in this thread: http://forum.pfsense.org/index.php/topic,28599.0.html.

    This corruption problem has since been fixed by upgrading PFSense to the latest version a couple days ago as I described earlier but the load balancing is still not working. I don’t dare try to remove the Default Gateway designation again for fear of corrupting the WAN port like before.

    Both of the WANs that use PPoE are working individually. I can use the ping tool under diagnostics and they both work, I can point traffic through each of them individually. But even though I added the Load Balancing Routing Group to the firewall rules the internet traffic will only go over the default gateway.

    Any advice would be greatly appreciated.

    Thanks in advance.
    Thomas



  • Post your config.xml file.
    Also, don't be afraid to change settings in the gateways now. That code is much more stable and shouldn't cause any more problems.

    GB



  • Hi

    Thanks for the reply, attached is the config.xml file.

    FYI: It would not allow me to upload a .xml file so I converted to .txt

    Also so you don't get confused when you look at the config file.  The Load Balancing group is setup for Lan1 which is labeled lan22…

    I also I have a third DSL dedicated to VoIP traffic which is on Lan2 labeled Lan11 this is working as it should.

    The problem is with the load balancing group of Wan1 and Wan2 as described in my first post for Lan1.

    Thanks

    config.txt



  • @gnhb:

    Post your config.xml file.
    Also, don't be afraid to change settings in the gateways now. That code is much more stable and shouldn't cause any more problems.

    GB

    hi i've recently reverted back to late august release because Load Balancing, particularly Gateway configuration seem to be problematic the whole of September and early October.

    Is Load Balancing/Gateway config ok now? (Oct 11 onwards)



  • Load Balancing should be fixed now. Please test again with a snapshot after Oct 10 13:00.



  • Great… I will test it today and let you know.

    Thanks



  • Yes, the problem is now fixed. The load balancing is now working correctly.



  • @gnhb:

    Load Balancing should be fixed now. Please test again with a snapshot after Oct 10 13:00.

    Updated to Oct 13 build. Indeed, Load Balancing works again. Thanks :)



  • HI,

    I'm using pfSense 2.0BETA4 from 30.11.2010.

    I have got  two WAN ports in Load Balancing. If I reboot pfSense, Load Balancing works fine, but if I disconnect WAN1 (rl0) and reconnect it after a short time, it comes up but there isn't any loadbalancing anymore.

    Syslog ist telling me this:

    Dec 1 16:19:04 	dnsmasq[1806]: using nameserver 8.8.8.8#53
    Dec 1 16:19:04 	dnsmasq[1806]: using nameserver 194.25.2.129#53
    Dec 1 16:19:04 	dnsmasq[1806]: reading /etc/resolv.conf
    Dec 1 16:17:47 	apinger: Starting Alarm Pinger, apinger(17178)
    Dec 1 16:17:47 	apinger: Exiting on signal 15.
    Dec 1 16:17:46 	dhclient[55051]: bound to 192.168.1.173 -- renewal in 432000 seconds.
    Dec 1 16:17:46 	check_reload_status: Rewriting resolv.conf
    Dec 1 16:17:46 	php: : rc.newwanip: on (IP address: 192.168.1.173) (interface: wan) (real interface: rl0).
    Dec 1 16:17:46 	dhclient: Creating resolv.conf
    Dec 1 16:17:46 	php: : rc.newwanip: Informational is starting rl0.
    Dec 1 16:17:46 	dhclient: Adding new routes to interface: rl0
    Dec 1 16:17:46 	check_reload_status: rc.newwanip starting rl0
    Dec 1 16:17:46 	dhclient: New Routers (rl0): 192.168.1.1
    Dec 1 16:17:46 	dhclient: New Broadcast Address (rl0): 192.168.1.255
    Dec 1 16:17:46 	dhclient: New Subnet Mask (rl0): 255.255.255.0
    Dec 1 16:17:46 	dhclient: New IP Address (rl0): 192.168.1.173
    Dec 1 16:17:46 	dhclient: ifconfig rl0 inet 192.168.1.173 netmask 255.255.255.0 broadcast 192.168.1.255
    Dec 1 16:17:46 	dhclient: Starting add_new_address()
    Dec 1 16:17:46 	dhclient: REBOOT
    Dec 1 16:17:46 	dhclient[55051]: DHCPACK from 192.168.1.1
    Dec 1 16:17:46 	dhclient[55051]: send_packet: No buffer space available
    Dec 1 16:17:46 	dhclient[55051]: DHCPREQUEST on rl0 to 255.255.255.255 port 67
    Dec 1 16:17:46 	dhclient[55051]: send_packet: No buffer space available
    Dec 1 16:17:44 	dhclient[55051]: DHCPREQUEST on rl0 to 255.255.255.255 port 67
    Dec 1 16:17:44 	dhclient: PREINIT
    Dec 1 16:17:43 	dhclient[19298]: exiting.
    Dec 1 16:17:43 	dhclient[19298]: exiting.
    Dec 1 16:17:43 	dhclient[19298]: short write: wanted 12 got 0 bytes
    Dec 1 16:17:43 	dhclient[19298]: short write: wanted 12 got 0 bytes
    Dec 1 16:17:43 	dhclient: FAIL
    Dec 1 16:17:43 	dhclient[55870]: My address (192.168.1.173) was deleted, dhclient exiting
    Dec 1 16:17:43 	dhclient[55870]: My address (192.168.1.173) was deleted, dhclient exiting
    Dec 1 16:17:43 	php: : HOTPLUG: Configuring interface wan
    Dec 1 16:17:43 	php: : DEVD Ethernet attached event for wan
    Dec 1 16:17:42 	kernel: rl0: link state changed to UP
    Dec 1 16:17:42 	check_reload_status: Linkup starting rl0
    Dec 1 16:16:54 	php: : MONITOR: WAN1 has high latency, removing from routing group
    Dec 1 16:16:54 	php: : MONITOR: WAN1 has high latency, removing from routing group
    Dec 1 16:16:53 	check_reload_status: reloading filter
    Dec 1 16:16:43 	apinger: ALARM: WAN1(192.168.1.1) *** down ***
    Dec 1 16:16:34 	kernel: rl0: link state changed to DOWN
    

    The Traffic Graph on startup page only shows traffic on WAN2 (rl1) but nothing on WAN1 (rl0).
    After a reboot, everything works fine again.

    I added my .XML config file but renamed it to .TXT.

    It would be fine, if you could help me.

    pfsense1.txt





  • Is this patch added in the Dec 10  00:14  i386 ?

    Because I still have got the problem, if WAN1 goes down and later comes up again, Load Balancing didn't work, only WAN2 is working.



  • YEah it is.
    Show me your system logs to see if your claim is right?



  • I will do on monday. Over the weekend I have got no access to my testing environment.



  • Hi,

    sorry for my late feedback but I didn't get it earlier. Today I updated to the latest snapshot, becaus I read in the rsc, that there were some fix for updating WAN interfaces IP adresse or something and now I just want to let you know, that it works for me again.

    Great job, thanks!


Locked