Re[2]: [Openh323gk-users] GkStatus bug

[Roman R. <ro...@rg...>]

Hello Zygmuntowicz,

Wednesday, May 31, 2006, 1:18:38 PM, you wrote:

It may be timeout or syncronisation issue. I try to repeat this
problem on my Windows GnuGK, but got another error:
RasTbl.cxx:
===
void RegistrationTable::PrintAllRegistrations(USocket *client, BOOL verbose)
{
    int i;
    char s[64],*buf;

    buf=new char[16384];
    buf[0]=0;
    for(i=0;i<512;i++){
        sprintf(s,"%03d 1234567890\r\n",i);
        strcat(buf,s);
    }
    client->TransmitData(buf);
    delete[] buf;
    PString msg("AllRegistrations\r\n");
    InternalPrint(client, verbose, &EndpointList, msg);
}
===

Result is:
.
.
510 1234567890
511 1234567890


Connection to host lost.

When I changed 512 to 800, I got:
.
.
601 1234567890
602 1234567890
603

Connection to host lost.

WSAGetLastError() was 10038 (WSAENOTSOCK) - very strange.
InternalPrint also sends data as one big message, so it may be same
problem.

ZM> Quite strange - it should not stop - we don't have any particular
ZM> limits for the status port output size.

ZM> ----- Original Message ----- 
ZM> From: "Roman Rusakov" <ro...@rg...>
ZM> Sent: Tuesday, May 30, 2006 9:56 AM


>> If I have ~200 endpoints, and do connect via telnet to GnuGK:
>> ===
>> OpenH323GK login: gkadmin
>> Password:
>> Version:
>> Gatekeeper(GNU) Version(2.2.4) 
>> Ext(pthreads=1,radius=1,mysql=1,pgsql=0,large_fdset=8192)
>> Build(Apr 20 2006, 06:51:11) Sys(Linux i686 2.4.20-8)
>> GkStatus: Version(2.0) Ext()
>> Toolkit: Version(1.0) Ext(basic)
>> Startup: Mon, 29 May 2006 17:58:04 GMT   Running: 0 days 13:51:39
>> ;
>> trace 0
>>
>> Output trace level is 0
>> ?
>> ...
>> ===
>> I receive 8*1518 bytes of info about registered endpoints and output
>> is hang :(
>> To continue I must send any command to GK via telnet.
>> Why GnuGK stop output after 8 TCP packets of data?
>>
>> -- 
>> Best regards,
>> Roman                          mailto:ro...@rg...

-- 
Best regards,
 Roman                            mailto:ro...@rg...