Bugs item #3511424, was opened at 2012-03-26 06:36
Message generated for change (Comment added) made by kargor
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=843359&aid=3511424&group_id=167540

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Network Layer
Group: None
Status: Open
Resolution: None
Priority: 7
Private: No
Submitted By: Stefan Rossbach (kargor)
Assigned to: Nobody/Anonymous (nobody)
Summary: Transport Layer establishes wrong connections

Initial Comment:
DEBUG 2012-03-25 18:00:39,554 (Socks5Transport.java:299) [SOCKS5] stream is 
unidirectional. Trying to wrap bidirectional one.
DEBUG 2012-03-25 18:00:39,554 (Socks5Transport.java:228) [SOCKS5] response 
connection is mediated, too, and the server does not allow bidirectional 
connections. Wrapped session established.
DEBUG 2012-03-25 18:00:39,560 (DataTransferManager.java:465) Bytestream 
connection changed SOCKS5 (mediated)
DEBUG 2012-03-25 18:00:39,560 (BinaryChannelConnection.java:53) SOCKS5 
(mediated) [[email protected]/Saros] ReceiverThread started.
INFO  2012-03-25 18:00:39,686 (ByteStreamTransport.java:71) Received request to 
establish a IBB bytestream connection from [email protected]/Saros
DEBUG 2012-03-25 18:00:39,692 (DataTransferManager.java:465) Bytestream 
connection changed IBB
DEBUG 2012-03-25 18:00:39,693 (BinaryChannelConnection.java:53) IBB 
[[email protected]/Saros] ReceiverThread started.

The mediated connection got downgraded to IBB. Also can anyone confirm that 
saros-con.imp.fu-berlin.de does not support bidirectional Socks5 Mediated 
connections ?


----------------------------------------------------------------------

>Comment By: Stefan Rossbach (kargor)
Date: 2012-06-28 07:39

Message:
It is not hard to reach 10 KB/s. Even a single transfered character takes
about 400 - 600 Bytes.

Transfering approx. 3 MBytes (average project size) at ISDN speed reminds
me to my first days in the internet using Napster.

----------------------------------------------------------------------

Comment By: Franz Zieris (franzzieris)
Date: 2012-06-28 04:03

Message:
It's hard to follow your line of thought.

Does this problem only occur on machines that are reachable via a dozen
IPs, or does this scenario also apply for less exotic settings? In order to
make your point clear you should try to find a minimum scenario.

Do I get you right: The problem is, that there are scenarios in which the
best possible connection between two Saros users is not established because
of the unfortunate execution order of the setup steps? And the only result
is a low-bandwidth connection? 
What are the "visible" results in a Saros Session (I imagine that even a
olympic typist can't beat 10 KByte/s)?

----------------------------------------------------------------------

Comment By: Stefan Rossbach (kargor)
Date: 2012-06-27 13:38

Message:
I can give you one:

Client A:  2 IP addresses (public, eth0) both not reachable

Client B: 8 IP addresses (public is reachable)

Smack impl. sequentielly try connecting to each IP, 2 seconds timeout, at
least it returns a mediated connection (if there is any)

Let the fun begin:

B should connect to A
in the meantime, A connects to B

B will get a mediated connection

B waits for possible timeout of A (which will be 2 * 8 = 16 sec.)
B timed out (hardcoded in Socks5Transport class and is less than 16
seconds), but A is still connecting.

A establish a direct connection, but B already checks if the mediated
connection
is bidirectional.

No response from A, because A have a direct connection.

Result:

Established connection will be IBB.

Solution, A connects to B, if that fails, B tries to connect to A. In that
order but not concurrently !

----------------------------------------------------------------------

Comment By: Franz Zieris (franzzieris)
Date: 2012-05-23 09:11

Message:
I can not evaluate the necessity of priority "7" on this. Could someone
explain it to me, please?

----------------------------------------------------------------------

Comment By: Norman Warnatsch (nwarnatsch)
Date: 2012-03-26 07:28

Message:
I don't think, that's a problem of saros-con.imp.fu-berlin.de. 

Sock5 (mediated) have to work like the following scenario:

Our server just sends requests from Requester to Target. Requester have to
know the address of the host and sends a negotiation with host-address to
the target. Target opens a tcp-connection to host, requester to host, too..
The host is now the proxy for the connection between Requester and Target.

Take a look to:
http://xmpp.org/extensions/xep-0065.html

So, saros-con.imp.fu-berlin.de  shouldn't have a problem with bidirectional
connections because it just should send the Sock5(mediated)-negotiations,
nothing else.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=843359&aid=3511424&group_id=167540

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Dpp-robot mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dpp-robot

Reply via email to