AW: BufferException with large messages

["Braumann Andreas" <a.braumann@xxxxxxxxxxxxxx>]

Hi!

Thanks for the response. We identified the problem and it was independent from SNACC. It became apparent because we used the wrong send and recv methods from ACE framework. Specifically, we used send and recv instead of send_n and recv_n and the large messages probably got split with the first ones and we didn't take that into account. So we tried to decode only a part of the message, and thus the Exception.

Nevertheless thanks for this piece of software, it helped us tremendously.

> -----Ursprüngliche Nachricht-----
> Von: Abarbanel, David (US SSA) [mailto:david.abarbanel@xxxxxxxxxxxxxx]
> Gesendet: Freitag, 18. August 2006 18:58
> An: Braumann Andreas; imc-snacc@xxxxxxx
> Betreff: RE: BufferException with large messages
> 
> I looked into this and unfortunately have been unable to reproduce the
> problem.  I assume you are using esnacc v1.7 with the 1.7.3 patch?
> Could you provide a sample ASN.1 definition that has hit this problem
> while data is being decoded.
> 
> Thanks,
> 
> Dave Abarbanel
> BAE Systems
> 
> -----Original Message-----
> From: owner-imc-snacc@xxxxxxxxxxxx [mailto:owner-imc-snacc@xxxxxxxxxxxx]
> On Behalf Of Braumann Andreas
> Sent: Wednesday, August 09, 2006 8:28 AM
> To: imc-snacc@xxxxxxx
> Subject: BufferException with large messages
> 
> 
> Hi there,
> 
> we stumbled upon a problem in the asn-buf.cpp file whereas a
> BufferException is raised in line 371 if we send very large ASN.1
> messages (several KB). This problem doesn't seem to appear with small
> messages.
> 
> The exception is raised more often while calling BDec on the buffer but
> sometimes on the PutSegRvs call on the buffer.
> 
> We found that the problem is not always reproducible, but the larger the
> messages, the more often it occurs (up to the point where it appears
> with almost every message). With messages in the 100s bytes size, it
> appears approximately 50% of the time, in the 1000s bytes size 90% of
> the time, in the 10s bytes size it didn't occur yet.
> 
> We are wondering if we are the only ones having these problems. Is this
> a bug in the buffer? We tried to include the fix from July 7, 2005, but
> to no avail.
> 
> What's interesting is that when we catch the BufferException and don't
> do anything with it everything seems to continue normally (message is
> successfully decoded and so on), so currently we catch all those
> Exceptions and don't handle them and it works. Still it would be nice to
> get some insights.
> 
> Andreas Braumann
> Synergis