Ian is the author of the unix socket stuff. Goodness knows he and I  
have tried to ensure it works correctly under heavy load.

a) All unix variations have slightly different behaviour under load.  I  
think Ian has tested on os-x and NetBSD.
b) Try disabling the interval timer  -notimer and see if that changes  
things?
c) Having a repeatable Sunit will greatly help debugging?
d) You must build a debuggable VM which dumps sufficient state about  
the status of the socket to understand why the
socket is being closed.
e) As pointed out they issue could be on the other side with you  
violating some rule and they close you down.

Please keep in mind this is not a language issue. The interesting and  
troublesome parts of the Socket code are just plain C code. The same  
issue could be occurring if the application was written in C or a host  
of other languages.


On Friday, September 12, 2003, at 08:43  AM, ducasse wrote:

>> From: Lukas Renggli <renggli at student.unibe.ch>
>> Date: Ven sep 12, 2003  17:18:34 Europe/Zurich
>> To: squeak-dev at lists.squeakfoundation.org
>> Subject: Socket Problems
>> Reply-To: The general-purpose Squeak developers list  
>> <squeak-dev at lists.squeakfoundation.org>
>>
>> Hi,
>>
>> we've got a production server for an insurance company running a  
>> Squeak
>> 3.5 image on the 3.6b6 VM on Debian Linux. Unfortunately there are
>> mysterious problems with the sockets when accessing PostgreSQL from
>> within Squeak. We wrote workarounds, that just retry if a connection  
>> is
>> failing, but this isn't helping much anymore.--
======================================================================== 
===
John M. McIntosh <johnmci at smalltalkconsulting.com> 1-800-477-2659
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
======================================================================== 
===