Hello, I use an SRV record as the proxy value for a sofia gateway. This particular record is configured for failover rather than load balancing. In other words, there is only one record per priority. I've verified Freeswitch will advance to the second host if the first becomes unavailable. It seems Freeswitch waits around 10 minutes before it tries the first host again. Is this delay a configurable value? Is there some kind of internal blacklist the failed host occupies before this timer expires? Is this list manageable from the CLI? Besides a 408, are there other response codes that will land a host on this apparent blacklist? - Jeff
Not really the answer, but... There's a gateway param called "ping", e.g: <!--send an options ping every x seconds, failure will unregister and/or mark it down--> <param name="ping" value="25"/> Then if it's marked down (via 202 or 404 perhaps<http://wiki.freeswitch.org/wiki/Sofia.conf.xml>), any bridge string referencing that gateway will know it's down before contact it and move on. (with error code NETWORK_OUT_OF_ORDER) You can see that showing up in the logs, and I think in the gateway status. I don't see a command to manually re-ping the gateway. -Avi On Mon, Feb 4, 2013 at 6:52 PM, Jeff Pyle <jpyle@fidelityvoice.com> wrote:
Hello,
I use an SRV record as the proxy value for a sofia gateway. This particular record is configured for failover rather than load balancing. In other words, there is only one record per priority. I've verified Freeswitch will advance to the second host if the first becomes unavailable.
It seems Freeswitch waits around 10 minutes before it tries the first host again. Is this delay a configurable value?
Is there some kind of internal blacklist the failed host occupies before this timer expires? Is this list manageable from the CLI?
Besides a 408, are there other response codes that will land a host on this apparent blacklist?
- Jeff
_________________________________________________________________________ Professional FreeSWITCH Consulting Services: consulting@freeswitch.org http://www.freeswitchsolutions.com
FreeSWITCH-powered IP PBX: The CudaTel Communication Server http://www.cudatel.com
Official FreeSWITCH Sites http://www.freeswitch.org http://wiki.freeswitch.org http://www.cluecon.com
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Hello, to solve it temporarily (I also don't know of a manual ping), you can ofcourse kill the gateway "sofia ... killgw" and bring it back up using a rescan of the profile ("sofia profile .... rescan reloadxml"). I suppose that the options-ping, like suggested below should work. I don't have any problem as far as I can say and I did quite some failover-tests. Kind regards, Michel On Mon, Feb 4, 2013 at 6:11 PM, Avi Marcus <avi@avimarcus.net> wrote:
Not really the answer, but...
There's a gateway param called "ping", e.g:
<!--send an options ping every x seconds, failure will unregister and/or mark it down--> <param name="ping" value="25"/>
Then if it's marked down (via 202 or 404 perhaps<http://wiki.freeswitch.org/wiki/Sofia.conf.xml>), any bridge string referencing that gateway will know it's down before contact it and move on. (with error code NETWORK_OUT_OF_ORDER)
You can see that showing up in the logs, and I think in the gateway status. I don't see a command to manually re-ping the gateway.
-Avi
On Mon, Feb 4, 2013 at 6:52 PM, Jeff Pyle <jpyle@fidelityvoice.com> wrote:
Hello,
I use an SRV record as the proxy value for a sofia gateway. This particular record is configured for failover rather than load balancing. In other words, there is only one record per priority. I've verified Freeswitch will advance to the second host if the first becomes unavailable.
It seems Freeswitch waits around 10 minutes before it tries the first host again. Is this delay a configurable value?
Is there some kind of internal blacklist the failed host occupies before this timer expires? Is this list manageable from the CLI?
Besides a 408, are there other response codes that will land a host on this apparent blacklist?
- Jeff
_________________________________________________________________________ Professional FreeSWITCH Consulting Services: consulting@freeswitch.org http://www.freeswitchsolutions.com
FreeSWITCH-powered IP PBX: The CudaTel Communication Server http://www.cudatel.com
Official FreeSWITCH Sites http://www.freeswitch.org http://wiki.freeswitch.org http://www.cluecon.com
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
_________________________________________________________________________ Professional FreeSWITCH Consulting Services: consulting@freeswitch.org http://www.freeswitchsolutions.com
FreeSWITCH-powered IP PBX: The CudaTel Communication Server http://www.cudatel.com
Official FreeSWITCH Sites http://www.freeswitch.org http://wiki.freeswitch.org http://www.cluecon.com
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
participants (3)
-
Avi Marcus
-
Jeff Pyle
-
Michel Brabants