I encounter "The Dialogic System Service service terminated with service-specific error %%16383" when try to start the device.
I'm using D/600JCT-2E1 card in window 2008 server SP1. It's has 16G of memory.
I also notice the following error
MERCCONFIG - DLGCMPD driver failed to start.
DLGCSERV - RunProgsInKey() - CreateProcess FATAL ERROR : SctsAssi.exe -LSctsassi.log
DLGCSERV - RunProgsInKey() - SctsAssi.exe -LSctsassi.log returned fatal error.
Anyone can help?
The mains problem would have been:
DLGCSERV - RunProgsInKey() - CreateProcess FATAL ERROR : SctsAssi.exe -LSctsassi.log
This points issue towards windows short paths not being enabled on the system. The dialogic install process requires windows short paths to be enabled in order for some lower level services to run correctly on newer 64-bit O/S (I.e. 2008 R2 / 2012).
For more information refer to the following Microsoft technical note on setting short paths on windows system:
technet.microsoft.com/.../cc778996(v=ws.10).aspx
Jeff
Hi,
Are you are setting some sort of proxy settings as well in IP virtboard? What address is selected in DCM for RTP traffic (ie default IP address tab).
Jeff
Thanks for the quick response, Jeff. I don't think we're setting any proxy settings, our startup code is below.
When we first installed the system, the subnet mask on the VoIP lan was set incorrectly resulting in traffic being routed via the default gateway. When we realized this we corrected the subnet but the traffic still included the VIA header. We then uninstalled HMP, ran regclean.bat, and re-installed pointing to the 10.37.56.x interface. Now traffic is directed to the 10.37.56 adapter but still includes the VIA header referencing 192.168.10. Is it possible there is some remnant of the previous config that is not being cleared by the reinstall and regclean.bat? How can I troubleshoot this?
Startup code:
IPCCLIB_START_DATA ipcclibstart;
IP_VIRTBOARD ip_virtboard[1];
GC_START_STRUCT gclib_start;
CCLIB_START_STRUCT cclib_start[] = {
{"GC_ISDN_LIB",NULL},
{"GC_IPM_LIB",NULL},
{"GC_H3R_LIB", &ipcclibstart},
{"GC_DM3CC_LIB",NULL},
};
INIT_IPCCLIB_START_DATA(&ipcclibstart, 1, ip_virtboard);
INIT_IP_VIRTBOARD(&ip_virtboard[0]);
ip_virtboard[0].sup_serv_mask = IP_SUP_SERV_CALL_XFER;
ip_virtboard[0].sip_msginfo_mask = IP_SIP_MSGINFO_ENABLE;
ip_virtboard[0].localIP.u_ipaddr.ipv4 = inet_addr(szLocalIP.c_str()); //szLocalIP is the local PC address on 10.37.56.x
gclib_start.cclib_list = cclib_start;
gclib_start.num_cclibs = 4;
GC_CCLIB_STATUSALL cclib_status_all;
if (gc_Start(&gclib_start) != GC_SUCCESS) .....
In this case, I think its getting pulled off the "default gateway address" set on the system itself from the NIC. Thus as a test you can change it to the 10.37.56 address, then I suspect it should get correct.
Jeff
I have a system with 2 Ethernet adapters, 10.37.56.x (dedicated to VoIP) and 192.168.10.x The default gateway is on the 192.168.10 network. We've installed HMP 3.0 Windows SU349 and configured it to use the 10.37.56.x address.
When we attempt to register to the PBX (also on the 10.37.56.x subnet), the REGISTER packet is sent out the 10.37.56.x adapter but with a VIA header referencing the 192.168.10.x network. I cannot tell why this is.
I've confirmed with traceroute that traffic is directly routed out the 10.37.56.x network. I've also tried setting the local 10.37.56.x IP in the u_ipaddr.ipv4 member of the IP_VIRTBOARD structure, but still the VIA header is being inserted.
On what basis does HMP decide to insert VIA headers in outgoing SIP messages?
Do Media Toolkit Overlays Work In HMP4.1? There is a fair bit of official information and a few demos about getting the 'sm_Overlays' working. I've not been able to get any demo to actually work.We've tried seemingly every configuration, variation AND combination there is. (Like JPEG vs. YUV, etcetera.)
Nor have we been able to get the "6.2 Stream Manipulation Media Toolkit Example Code" from the Media Toolkit API v3 (August, 2011) to work. This includes a person who has worked on Dialogic for a very long time.
Interestingly, I can't find a single example (in English) of anyone on the entire internet that even talks about them, let alone having overlays actually working.
PS: I'm new here and I inadvertently posted this question in the wrong forum. I am unable to make any change to that. I'm not spamming. :-)
Thank you.
Hi,
This is not currently support right now on the Windows HMP 3.0 SU baseline. However, there are plans to add it in this year once tested. Stay tuned for 1st half of this year potentially.
Jeff
Can anyone tell me if HMP SU361 is supported on VMWare6.0 as a VMWare “Virtual machine version 11” running Windows 2012-R2?
I use HMP with ipm as bellow code:
char str[MAX_STRING_SIZE];
//---------inband mode--------
IPM_PARM_INFO parmInfo;
unsigned long ulParmValue = DTMFXFERMODE_INBAND;
parmInfo.eParm = PARMCH_DTMFXFERMODE;
parmInfo.pvParmValue = &ulParmValue;
ipm_SetParm(voice[index_voice].deviceInfo[IPM].devH, &parmInfo, EV_ASYNC);
But when i press key, I must wait 10s to received from my program. what is wrong? it's by MSC or how to setting in HMP.
This is my capture file pcap.
http://www.mediafire.com/download/wv6c7wu94j45676/0203_1.pcap
I have the same problem. How can build an HMP application on a linux 64 bits version (RedHat Enterprise Linux 7.2)?
I have the same issue. I have added the -m32 flag but the linker still tries to link against the 64 bits system libraries (for example /usr/lib64/libpthread). How do I solve this?
Does that mean that HMP doesn't support 64 bits (RedHat Enterprise Linux 7.1)?
Thanks.
Platform --> CentOS Release 6.3 64bits
HMP Package --> lnxHMP_4.1_129.tgz
Before I completed install HMP, it asked for 32-bits libstdc++.
Atfer install libstdc++ 32bits, HMP installation seem to be completed.
Test make demo in gc_basic_call_model the result is
cc gc_basic_call_model.o -L/usr/dialogic/lib -lgc -o gc_basic_call_model
/usr/bin/ld: skipping incompatible /usr/dialogic/lib/libgc.so when searching for -lgc
/usr/bin/ld: cannot find -lgc
Is this related to differentiate between 64bits platform and 32bits lib? and How to solve this problem?
Thanks.
The HMP 4.1 Su213 release has added support for RH 7.1, which SU are you using? If earlier version then you need to upgrade.
Also, all application must be built to run as 32-bit mode on 64 bit O/S. So you need to make sure you have all gcc/g++ and libstdc/libstdc++ packages installed for 32 bit development libs on the box. Since the 64-bit versions are not used.
Jeff
I think you need to be more specific on what exactly you mean by crash? was it a BSOD, or application level crash. If so, did you check the dmp file to see where the stack points towards.
Did you recompile your application with SU361 to match that of the SU installed when running on the system?
Jeff
Hi Jeff,
thank you for your response.
I have already compile my application with SU361.
I develop under Visual studio 12.
I have try to launch my application in debug mode in VS12, and the application has stopped in cheetah thread. I can't see the stack because i haven't the source of the cheetah Library.
Yvan
Ok, hard to say if the issue is HMP then, as we have not qualified against VS12 in this case. It could be that something changed and the libs need to be rebuilt for newer compilers. As you know the HMP libs are build on VC 8.0, so maybe some interop issue exists.
I know that seen testing on VS 2008, and VS 2010. Not aware is anyone has tried something newer and what the results were in that case.
Jeff
Hi, I use the DIVA SDK.
Actually I detect a busy line ( if have don't receive a tone 42 in the first 30 seconds ) I said the line busy.
What is a better way to detect a busy line.
Something like a fax. If I receive tone 89, it's a fax ??
Hi,
We successfully use HMP 3.0 SU 361 on Windows 7, 8, 10 and Server 2008 and 2012.
The applications are built under Visual Studio 2013.
It's strange that your application crashes on start. It might be an initialization issue.
Another problem that we have encountered was that some other ill-behaving applications would install "libeay32.dll" and "ssleay32.dll" in the System32 folder, and the version of those 2 dlls was not compatible with the ones which HMP uses, and which are delivered in HMP/lib folder. The problem is that when Dialogic dlls are loaded, they find the ssl dlls inthe System32 folder first, and so Dialogic dlls fail to load, possibly leading to a crash, depending on how they are linked into your own app.
Chris
Hi,
I am facing a problem sometimes when using fx_sendfax in SIP enviroment .
Usualy it works fine , i conect ipt and fax device when the call is connected and issues fx_sendfax (g711 fax ) .
Sometimes i get this error in log:
ssp_x86Linux_boot: QERROR_WARNING(2) qkernerr.h 41400 qkernerr.h 41003
ssp_x86Linux_boot: QERROR_WARNING(2) qkernerr.h 41400 qkernerr.h 41003
ssp_x86Linux_boot:
Fssp_x86Linux_boot:
ssp_x86Linux_boot: #011Error Level : WARNING
ssp_x86Linux_boot: #011Error Level : WARNING
ssp_x86Linux_boot: #011Location Tag: 41400, Location File: qkernerr.h
ssp_x86Linux_boot: #011Location Tag: 41400, Location File: qkernerr.h
ssp_x86Linux_boot: #011Cause-Tag : 41003, Error File :qkernerr.h
ssp_x86Linux_boot: #011Cause-Tag : 41003, Error File :qkernerr.h
ssp_x86Linux_boot:
ssp_x86Linux_boot:
ssp_x86Linux_boot: GSTRM_Open[24]: qGStreamOpen returned 2 errno 0x41003
ssp_x86Linux_boot: GSTRM_Open[24]: qGStreamOpen returned 2 errno 0x41003
ssp_x86Linux_boot: CFile::Open: ERROR: GSTRM_Open(read)
ssp_x86Linux_boot: CFile::Open: ERROR: GSTRM_Open(read)
ssp_x86Linux_boot: CStrmCache::Open: ERROR: FILE_Open()
ssp_x86Linux_boot: CStrmCache::Open: ERROR: FILE_Open()
ssp_x86Linux_boot: [24]T30::ExecFromHostIdleState: PAGE_Open error
ssp_x86Linux_boot: [24]T30::ExecFromHostIdleState: PAGE_Open error
ssp_x86Linux_boot: QFC3[24]: processQFC2Msg(SESSION_STARTING_CHILD): Got QFC2_MsgSessionCmplt, changing QFC3 state to Allocated
ssp_x86Linux_boot: QFC3[24]: processQFC2Msg(SESSION_STARTING_CHILD): Got QFC2_MsgSessionCmplt, changing QFC3 state to Allocated
I have no idea what this means .
None of the fax functions returned any error and tiff exists on its location .
/* I connected ipt and fax channels . */
value = DFC_AUTO;
ret = fx_setparm(pfax->Faxdev, FC_SENDCONT, (void *)&value);
if ((fhandle = open(filename, O_RDONLY)) == -1) {
return -1 ;
}
idx=0;
memset( df_iott[idx], 0, sizeof( DF_IOTT ) );
fx_setiott(df_iott[idx], fhandle, DF_TIFF, DFC_AUTO);
df_iott[ idx ].io_type |= IO_CONT;
df_iott[ idx ].io_type |= IO_EOT;
/* Set initial state of fax channel */
if(fx_initstat(pfax->Faxdev, DF_TX) == -1) {
return -1 ;
}
if(fx_sendfax(pfax->Faxdev, pfax->df_iott, EV_ASYNC|DF_PHASEB|DF_PHASED) == -1) {
return -1 ;
}
I always send one tiff file sometimes it has one and sometimes multiple pages