Hi,
it's a clean approach that costs performance. But it should be tested first,
how much performance it costs.
Now to this patch:
+ /* store temporary DLR data for SMSC ACK */
+ if (DLR_IS_ENABLED_DEVICE(msg->sms.dlr_mask) &&
!uuid_is_null(sms->sms.id)) {
+ Octstr *mid;
+ char id[UUID_STR_LEN + 1];
+
+ uuid_unparse(sms->sms.id, id);
+ mid = octstr_create(id);
+
+ dlr_add(conn->id, mid, sms);
dlr_add without dlr_find is not good. It will just grow DB...
+ if (conn->reroute_to_smsc) {
+ sms->sms.smsc_id = octstr_duplicate(conn->reroute_to_smsc);
+ } else {
+ sms->sms.smsc_id = octstr_duplicate(conn->id);
+ }
reroute belongs to abstraction layer...
Thanks,
Alexander Malysh
Am 28.07.2010 um 20:59 schrieb Rene Kluwen:
> No, you are right. I saw your solution only after I made the
> smsbox-route-between patch.
> I think your solution is cleaner. It depends of the routing options that are
> already available.
> I attached your (Stipe's) patch another time, as I see that it is still not
> committed to svn trunk.
>
> So the 2 questions here:
>
> 1. What do the other developers thing about this patch,
>
> and:
>
> 2. Can this patch (either one) submitted to Kannel svn, so people can
> benefit from smsbox (open smppbox) to smsbox routing?
>
> Previously, there hasn't been a lot of requests for it, other than be able
> to use sms-service from an smsbox request.
> Currently, I have been getting a lot of requests for this functionality
> because open smppbox depends on it.
>
> == Rene
>
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf
> Of Stipe Tolj
> Sent: Wednesday, 28 July, 2010 17:44
> Cc: [email protected]
> Subject: Re: [PATCH] SMS Box routing
>
>> As per popular request, I made a patch for bearerbox that enable the
>> following configuration:
>>
>> # multi-group:
>> group = smsbox-route-between
>> smsbox-id1 = localbox1
>> smsbox-id2 = remotebox1
>> # shortcode is optional
>> shortcode = +1234567;+7654321
>>
>>
>> This is the opposite of "reroute" for smsc's. It routes messages from
>> smsbox/sqlbox/open smppbox to another box that is connected.
>>
>> This allows for:
>>
>> 1. client-to-client routing, concerning open smppbox.
>> 2. Call sms services from one of the boxes. Right now you can only call
> them
>> via an smsc.
>> 3. Many more interesting things
>
> thanks a lot Rene for the patch.
>
> I'm at the moment -0 for this, due to the fact that it logically duplicates
> the
> behavior that we already have via the 'smsc = loopback' type.
>
> In your approach you do the re-routing here in the bearebox abstractive
> layer,
> where the loopback smsc module does it in the smsc module layer. The
> benefits
> here are we get the MT and MO entry logged to access-log while running, and
> there are no further config groups needed, as you can define the same
> routing
> you intend between smsbox instances bound to the same bearerbox by using the
> loopback smsc, and the smsbox-route group already.
>
> Do I miss some advantages here from this approach?
>
> Stipe
>
> --
> -------------------------------------------------------------------
> Kölner Landstrasse 419
> 40589 Düsseldorf, NRW, Germany
>
> tolj.org system architecture Kannel Software Foundation (KSF)
> http://www.tolj.org/ http://www.kannel.org/
>
> mailto:st_{at}_tolj.org mailto:stolj_{at}_kannel.org
> -------------------------------------------------------------------
>
>
> <smsc_loopback.diff>
