On Wed, May 4, 2011 at 11:49 PM, George Niculae <[email protected]> wrote:

>
> On Wed, May 4, 2011 at 6:39 PM, Irena Dolovčak 
> <[email protected]>wrote:
>
>> Hi,
>>
>> we have a problem with an high-availability sipxecs installation. we are
>> trying to create a sipxecs cluster containing two sipxecs (4.4.0) servers,
>> each of them having their own sipxbridge. while we are successful in
>> creating the cluster itself, we are unable to dedicate each of the two
>> trunks to the appropriate sipxbridges.
>>
>> our goal is to have one sip trunk from one provider, and another one from
>> another provider. if one server would fail, all of our phones would have
>> another way to communicate with the world. first we created sip trunk for
>> the first provider and assigned it to sipxbridge-1. then we enabled trunking
>> server role on the redundant server, and created the second sip trunk,
>> assigning it to the sipxbridge-2. although it gave us the choice of choosing
>> between sipxbridge-1 and sipxbridge-2 when we were creating the second sip
>> trunk, it didn't  apply our configuration regarding on which sipxbridge
>> should the trunk reside.
>>
>> we could see that second trunk wasn't applied to the sipxbridge-2 in three
>> different ways:
>>
>> 1) when we would go to the
>> DEVICES->GATWAYS->NAME_OF_THE_GATEWAY->CONFIGURATION, under 'Built-in
>> SipTrunk SBC' there would still be 'sipxbridge-1', instead of 'sipxbridge-2'
>>
>> 2) Since our trunk should register with the ITSP, under 'SIP Trunk SBC
>> Statistics' in DIAGNOSTICS it was clearly shown that our trunk has
>> registered with the ITSP while it was on the first server, not the second
>> one.
>>
>> 3) when we crashed the first server to see if second will make the calls,
>> our phones couldn't call external (not sipxecs registered) numbers, while it
>> could call all the extensions that were registered with either of the
>> sipxecs servers. while going through the logs to see the problem, it was
>> obvious that secondary sipxecs didn't know what to do with them, since he
>> didn't have those numbers as registered, and since he didn't have a bridge
>> to send those calls to.
>>
>> we tried to do the same configuration with 4.2.1 sipxecs cluster. all the
>> configuration was the same, the only difference was that we've installed
>> older version of sipxecs, and it worked. all three mentioned ways to check
>> if the sip trunk was assigned to the right bridge were  ok. also, when we
>> tried making a call to the outside world while the first server was down,
>> that call was successful.
>>
>> therefore, we think this is a bug in the new sipxecs version.
>>
>> Should we open a new jira for the problem?
>>
>>
> Hi Irena,
>
> yes, please open up a Jira and provide logs on debug
>
>
Looks like sipxbridge.xml generated on secondary is somehow messed up -
sipconnect.sipgate.de itsp-account does not have user name / password,
however it looks good on the primary, could you verify if copying
itsp-account entry for sipconnect.sipgate.de from primary to secondary
solves this issue? Check etc/sipxpbx/sipxbridge.xml

<itsp-account>
    <itsp-proxy-domain>sipconnect.sipgate.de</itsp-proxy-domain>
    <sipxecs-lineids>
    <sipxecs-lineid>13</sipxecs-lineid>
    </sipxecs-lineids>
    <user-name>ELIDED</user-name>
...
</itsp-account>

Thanks,
George
_______________________________________________
sipx-users mailing list
[email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-users/

Reply via email to