[JDEV] ICQ Transport Weirdness

Schuyler Heath sheath at jabber.org
Wed May 30 23:53:50 CDT 2001


Remove the <tcp/> section from your config so that peer 2 peer TCP is disabled.  The peer code in CVS
is very very unstable.

Schuyler

On Wed, May 30, 2001 at 09:53:50AM -0400, Benjamin Reed wrote:
> I'm not sure what caused this, but lately ICQ transport has been
> flaking out on me a lot.  The log file makes no sense, but is
> consistent.  Every time it dies, I end up with something like this:
> 
> ---(snip!)---
> Tue May 29 22:08:09 2001  deliver.c:649 delivering to instance 'icq.opennms.dyndns.org'
> Tue May 29 22:08:09 2001  mtq 80C4438 entering from pth
> Tue May 29 22:08:09 2001  mtq 80C4438 queue call 81D48B8
> Tue May 29 22:08:09 2001  icq_message_out.c:66 New message, 1 parts
> Tue May 29 22:08:09 2001  mtq 80C4438 leaving to pth
> Tue May 29 22:08:09 2001  mio.c:515 calling the connect handler for mio object 81FAC50
> Tue May 29 22:08:09 2001  peer.c:285 Session[ben at opennms.dyndns.org], sending peer init
> Tue May 29 22:08:09 2001  tcp.c:170 TCP INIT Len 28
> Tue May 29 22:08:09 2001  tcp.c:82 New packet 28
> Tue May 29 22:08:09 2001  tcp.c:32 tcp packet error: 28 26
> Tue May 29 22:08:09 2001  dnsrv.c:143 dnsrv: Read error on coprocess(1): 0 Success
> Tue May 29 22:08:09 2001  dnsrv.c:157 DNSRV CHILD: out of loop.. exiting normal
> ---(snip!)---
> 
> My ICQ transport is running in a separate process; it's been
> working for a good month up until about a week ago.  Here's the
> icqtrans.xml file I use to start it up:
> 
> ---(snip!)---
> <jabber>
>  <service id="icqlinker">
>   <uplink/>
>   <connect>
>    <ip>127.0.0.1</ip>
>    <port>5234</port>
>    <secret>nice try!</secret>
>   </connect>
>  </service>
>  <service id="icq.opennms.dyndns.org">
>   <icqtrans xmlns="jabber:config:icqtrans">
>    <instructions>Please enter your ICQ number (in the "username" field), nickname,
>     and password.  Leave the "username" field blank to create a new ICQ number.</instructions>
>  
>    <search>Search for ICQ users</search>
>  
>    <vCard>
>     <FN>ICQ Transport</FN>
>     <DESC>This is ICQ Transport</DESC>
>     <URL>http://opennms.dyndns.org/</URL>
>    </vCard>
>  
>    <!-- This should be a prime number close to the amount of concurrent users you expect to have. -->
>    <prime>501</prime>
>  
>    <!-- enables full TCP support -->
>    <tcp><ports/></tcp>
>  
>    <!--   Use the <ports/> to control the port range ICQ will use to listen for
>           incomming TCP connections.  If the ports section is not present,
>           ICQ-t will not listen on any port and make outgoing TCP connections only.
>           Remove the TCP section to disable TCP completely -->
>    <tcp>
>  
>     <ports>
>      <min>2000</min>
>      <max>3000</max>
>     </ports>
>    </tcp>
>     -->
>  
>    <dnsrv>
>      <host>icq.mirabilis.com</host>
>      <id>icq.dnsrv</id>
>      <delay>300</delay>
>    </dnsrv>
> 
>   </icqtrans>
>  
>   <load>
>    <icqtrans>./icq-cvs/icqtrans.so</icqtrans>
>   </load>
>  
>  </service>
>  
>  <service id="icq.dnsrv">
>   <load>
>    <dnsrv>./dnsrv/dnsrv.so</dnsrv>
>   </load>
>   <dnsrv xmlns="jabber:config:dnsrv">
>    <resend>icq.opennms.dyndns.org</resend>
>    <cachetimeout>300</cachetimeout>
>   </dnsrv>
>  </service>
> </jabber>
> ---(snip!)---
> 
> ...it's running on the current CVS version of ICQTrans, but
> switching around on versions doesn't seem to help.  When I
> restart it, everything works again, but it's dead by the next
> day.  Any ideas what's going on?
> 
> -- 
> Ben Reed (ben at opennms.org)
> http://www.opennms.org/
> "That's gotta be a JDK issue.  I can tell by the stack pointer." -- Shane
> _______________________________________________
> jdev mailing list
> jdev at jabber.org
> http://mailman.jabber.org/listinfo/jdev
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
URL: <https://www.jabber.org/jdev/attachments/20010530/68f08cae/attachment-0002.pgp>


More information about the JDev mailing list