hc-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Becke <be...@u.washington.edu>
Subject Re: [ATTN: Mike Becke] Baikanur, we have a problem
Date Thu, 24 Apr 2003 21:38:26 GMT
Agreed.  This may be the only way to go.  I'll try a few more things 
before I give up though.

Mike

Oleg Kalnichevski wrote:
> Mike,
> 
> The worst case scenario would be to test what platform we are running on
> and to perform state SSL connections checks on newer JVMs only (>=
> 1.4.0). I would rather openly admit and document the fact that SSL
> connection pooling works properly on JVM 1.4.1 only than keep on piling
> up messy code on top of messy code  
> 
> Oleg
> 
> 
> 
> On Thu, 2003-04-24 at 23:08, Michael Becke wrote:
> 
>>It appears to be more complicated than that.  For me, it seems that any 
>>attempt to read from an SSL connection (tunneled or not) before it has 
>>been written to causes the connection to be unusable.  This seems to 
>>apply to 1.3 and 1.2.
>>
>>For example, if we change the following in HttpConnection.isStale():
>>
>>         int byteRead = inputStream.read();
>>         if (byteRead == -1) {
>>             // again - if the socket is reporting all data read,
>>             // probably stale
>>//                            isStale = true;
>>         } else {
>>             inputStream.unread(byteRead);
>>         }
>>
>>Notice that reading a -1 no longer causes the connection to be stale. 
>>This change, though keeping the connection from being closed, caues the 
>>connection to not work for any subsequent method.
>>
>>If anyone has any ideas I would be more that happy to hear them. I will 
>>play around some more and see what I can figure out.
>>
>>Mike
> 
> 
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: commons-httpclient-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: commons-httpclient-dev-help@jakarta.apache.org
> 


Mime
View raw message