or use metadata to store it... Alex
Am 27.10.2011 um 10:36 schrieb Aarno Syvänen: > Hi Alex, > > yes, I agree that a new dlr field would be better. > > Aarno > > On 27.10.2011, at 09:51, Alexander Malysh wrote: > >> Hi Aarno, >> >> dict is not good because bearerbox can be restarted in the meantime and if >> you have much traffic >> dict will fill the memory. Maybe additional field in DLR DB? >> >> Alex >> >> Am 27.10.2011 um 09:43 schrieb Aarno Syvänen: >> >>> Hi, >>> >>> I work on patch that uses dict mapping incoming smpp client box id to msg >>> id- >>> When DLR is retuned, original box id is restored using this dict. >>> >>> MOs could be routed by smsc id, as presently, but by smppbox. >>> >>> Aarno >>> >>> On 27.10.2011, at 09:31, Alexander Malysh wrote: >>> >>>> Hi, >>>> >>>> I also don't like separate bearerbox connections but how should multiple >>>> smppbox setup handled if ESME can connect >>>> to each of them? Static routing in bearerbox doesn't work here. >>>> >>>> Alex >>>> >>>> Am 27.10.2011 um 09:06 schrieb Aarno Syvänen: >>>> >>>>> Hi, >>>>> >>>>> patch as I posted it is indeed not needed. But do we need separate >>>>> bearerbox connection for every smppbox client ? IMHO, this should >>>>> be changed, too. >>>>> >>>>> Aarno >>>>> >>>>> On 26.10.2011, at 01:06, Alexander Malysh wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> seems reasonable for me. Why we need the patch from Aarno then? >>>>>> >>>>>> Alex >>>>>> >>>>>> Am 21.10.2011 um 22:41 schrieb Rene Kluwen: >>>>>> >>>>>>> Yes, opensmppbox opens a separate box connection per connected client. >>>>>>> >>>>>>> I've heard before that using system-id is indeed more useful. So if you >>>>>>> want, go for that by using use-systemid-as-smsboxid or simply set the >>>>>>> system-type equal to system-id in clients.txt. >>>>>>> >>>>>>> -----Original Message----- >>>>>>> From: Alexander Malysh [mailto:[email protected]] On Behalf Of >>>>>>> Alexander Malysh >>>>>>> Sent: Thursday, 20 October, 2011 10:05 >>>>>>> To: Rene Kluwen >>>>>>> Cc: 'Aarno Syvänen'; [email protected] >>>>>>> Subject: Re: Using smppbox id >>>>>>> >>>>>>> But the box-id is for the whole box connection or does smppbox open >>>>>>> extra >>>>>>> box connection to bearerbox for >>>>>>> each ESME? >>>>>>> >>>>>>> Alex >>>>>>> >>>>>>> >>>>>>> Am 19.10.2011 um 23:22 schrieb Rene Kluwen: >>>>>>> >>>>>>>> Because the return messsges should be routed to the original client >>>>>>>> that >>>>>>> sent the first message to begin with... >>>>>>>> >>>>>>>> >>>>>>>> ----- Oorspronkelijk bericht ----- >>>>>>>> Van: Alexander Malysh <[email protected]> >>>>>>>> Verzonden: woensdag 19 oktober 2011 21:28 >>>>>>>> Aan: Rene Kluwen <[email protected]> >>>>>>>> CC: 'Aarno Syvänen' <[email protected]>; [email protected] >>>>>>>> Onderwerp: Re: Using smppbox id >>>>>>>> >>>>>>>> why box-id per client? I meant box-id per smppbox. Why do you want >>>>>>>> box-id >>>>>>> per client? >>>>>>>> >>>>>>>> Alex >>>>>>>> >>>>>>>> Am 19.10.2011 um 21:17 schrieb Rene Kluwen: >>>>>>>> >>>>>>>>> So you want a config option for boxc-id per client? >>>>>>>>> This is the same as configuring a system-type, isn't it? >>>>>>>>> >>>>>>>>> I agree, it's a hack. But better than cluttering the config files. >>>>>>>>> >>>>>>>>> == Rene >>>>>>>>> >>>>>>>>> -----Original Message----- >>>>>>>>> From: [email protected] [mailto:[email protected]] On >>>>>>> Behalf >>>>>>>>> Of Alexander Malysh >>>>>>>>> Sent: Wednesday, 19 October, 2011 19:41 >>>>>>>>> To: Aarno Syvänen >>>>>>>>> Cc: [email protected] Devel >>>>>>>>> Subject: Re: Using smppbox id >>>>>>>>> >>>>>>>>> Hi, >>>>>>>>> >>>>>>>>> just looked through the source code and I can only agree with Andreas: >>>>>>>>> system-type has nothing todo with box-id. >>>>>>>>> I don't know who uses it and why but it's totally wrong. I would just >>>>>>> remake >>>>>>>>> this part and make it straight forward >>>>>>>>> from design and understanding perspective: kill existing box-id hack >>>>>>>>> and >>>>>>>>> implement clean config options for box-id >>>>>>>>> and use _only_ these. >>>>>>>>> >>>>>>>>> @Aarno: you changed the code for data_sm only, how about submit_sm? >>>>>>>>> >>>>>>>>> Alex >>>>>>>>> >>>>>>>>> Am 19.10.2011 um 15:26 schrieb Aarno Syvänen: >>>>>>>>> >>>>>>>>>> This relies on assumption that no two smppbox share a client. I >>>>>>>>>> cannot >>>>>>>>> accept this. >>>>>>>>>> Besides, I can have two smppboxes connected to my application >>>>>>>>>> >>>>>>>>>> Aarno >>>>>>>>>> >>>>>>>>>> On 19.10.2011, at 15:19, Rene Kluwen wrote: >>>>>>>>>> >>>>>>>>>>> Like I said before: In clients.txt, you can put system-type to the >>>>>>>>>>> opensmppbox-id and you are all set. >>>>>>>>>>> >>>>>>>>>>> == Rene >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> -----Original Message----- >>>>>>>>>>> From: [email protected] [mailto:[email protected]] On >>>>>>>>> Behalf >>>>>>>>>>> Of Aarno Syvänen >>>>>>>>>>> Sent: Wednesday, 19 October, 2011 09:35 >>>>>>>>>>> To: [email protected] Devel >>>>>>>>>>> Subject: Re: Using smppbox id >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> In my case, there is an application between bearerbox and smppbox. >>>>>>>>>>> Thus application must route to smppbox and not to its clients. >>>>>>>>>>> >>>>>>>>>>> Aarno >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> On 18.10.201 >>>>>>>> >>>>>>>> >>>>>>>> [Het originele bericht is niet volledig opgenomen] >>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>> >>>> >>> >>> >>> >> >> > > >
