EFW Support

Support => General Support => Topic started by: Maestrale on Friday 28 May 2010, 06:58:05 pm



Title: Failover with EFW 2.4 is working now ?
Post by: Maestrale on Friday 28 May 2010, 06:58:05 pm
I have two uplink working and I've upgrade to 2.4
I  try to configure the failover  settings in the uplink config and, if I deactivate one uplink trough the dashboard\system\uplinks menu
failover is working or seems to be.... but, if I phisically disconnect the primary uplink wan cable, it doesent work, and after 20 seconds the secondary uplink go to DEAD state too ....???
somebody use failover or have try it ? if yes were I'm wrong ????


Title: Re: Failover with EFW 2.4 is working now ?
Post by: Maestrale on Friday 04 June 2010, 04:31:44 pm
Nobody have try if failover is working well ?


Title: Re: Failover with EFW 2.4 is working now ?
Post by: anvaruk on Saturday 05 June 2010, 02:44:06 am
have you tried adding, lets say ww.google.com into
"Check if these hosts are reachable"

in uplink editor ?


Title: Re: Failover with EFW 2.4 is working now ?
Post by: Maestrale on Monday 07 June 2010, 04:28:17 pm
Yes Google and opendns too ,
I've try to simulate a dead uplink disconnecting the telephon line instead of the wan cable too, but i have the same issue.
the secondary uplink go in dead state too.
all works perfect if I unceck the activate mark in the dashboard....after 5 / 10 sec. all works whith the secondary uplink when I Active it all traffic return on primary...
I don't undestand...


Title: Re: Failover with EFW 2.4 is working now ?
Post by: Maestrale on Monday 07 June 2010, 06:07:58 pm
Excuse me to all  :(
My fault ,all works perfect , I had put a wrong rules in the policy routing that prevented the secondary uplink to correct ping google and opendns
Sorry to all again....


Title: Re: Failover with EFW 2.4 is working now ?
Post by: jantram on Saturday 12 June 2010, 05:06:01 am
Interesting tidbit...

If you have have an interface or kernel routing entry pointing to the "dead" up link interface, the EFW kernel will direct the traffic out the dead port even though default route has been changed to the new uplink.

Entering ifconfig eth3 down removed the route from my table and allowed me to check the locally hosted domains.

This is likely obvious to people who would have this issue, but it is interesting to me, all the same.


Title: Re: Failover with EFW 2.4 is working now ?
Post by: jantram on Saturday 12 June 2010, 07:09:18 am
Followup to the tidbit I posted earlier: Policy Based Routing will fix that kind of thing for you, just take a surgical tack to it.