----- Original Message -----From: MadovskySent: Sunday, January 02, 2011 2:08 PMSubject: Re: [Freeswitch-users] FreeSWITCH & Heartbeat / Pacemaker - Please Help setting upI use HB/Pacemaker, tried to hack the original FS init script,but didn't succeed to work it well until now (when a FS fails, the failover fails also in Pacemaker.so I decided to create a cron task that ping/telnet/start/stop every FSFanck----- Original Message -----From: Avi MarcusSent: Sunday, January 02, 2011 4:43 AMSubject: Re: [Freeswitch-users] FreeSWITCH & Heartbeat / Pacemaker - Please Help setting upThanks, I noticed the -nonat already.What do you mean by running sofia recover from my init script? I don't understand how that would be triggered by heartbeat/pacemaker. Any explanatory docs on the issue would be nice, but everything I've seen was explaining exact parts and nothing ever seemed to explain the general syntax of everything.Does anyone have a working example that they can wikify and explain?Thanks,Avi
On Sun, Jan 2, 2011 at 4:06 AM, Seven Du <dujinfang@gmail.com> wrote:
the wiki page already said you can start the spare FS ahead to speed
switch over up.
it means if you use ip_nonlocal_bind and start the spare FS ahead, you
only need to run sofia recover in you init script without waiting for
the actually FS start process.
--
On Sun, Jan 2, 2011 at 7:58 AM, Steven Ayre <steveayre@gmail.com> wrote:
> 1) It seems freeswitch takes 8 seconds to restart. So if it crashes
> (unlikely) calls should immediately get shunted over to the other box. How
> do I set that up?
>
> Just a tip, I start fs with -nonat and find that makes it start faster.
>
> Steve on iPhone
> On 1 Jan 2011, at 20:20, Avi Marcus <Avi@aMarcus.com> wrote:
>
> Hi - I'm setting up 2 parallel computers for a HA setup. (low volume so far,
> no need to load balance.)
> I already have heartbeat able to grab the public IP if the first box goes
> down, but since I'm basically completely unfamiar with pacemaker, I'd like
> some help.
> I'd imagine I'm not the only one doing this, there should be no reason for
> me to recreate the wheel.
> Info should be on: http://wiki.freeswitch.org/wiki/Freeswitch_HA
> Specifically:
> 1) It seems freeswitch takes 8 seconds to restart. So if it crashes
> (unlikely) calls should immediately get shunted over to the other box. How
> do I set that up?
> 2) How do I automatically trigger a sofia recover? I know via commanline
> it's "fs_cli -x "sofia recover" but where do I tell pacemaker to do that?
> And I suppose a reload_xml before (maybe after?) is a good idea, too.
> Also, can I put <param name="track-calls" value="true"/> in a global file,
> or does it need to be in the actual profile files?
> And anything else I may be overlooking.
> Thanks guys!
> -Avi Marcus
>
> _______________________________________________
> 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
>
> _______________________________________________
> 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
>
>
About: http://about.me/dujinfang
Blog: http://www.dujinfang.com
Proj: http://www.freeswitch.org.cn
_______________________________________________
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
_______________________________________________
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