1995-08-24 - Re: SSL CHALLENGE: 0.11 borken?

Header Data

From: Piete Brooks <Piete.Brooks@cl.cam.ac.uk>
To: jered@MIT.EDU
Message Hash: 1d9f8e51ff1f0e6e4df91259c0d90ae547eadcfd1e7fa380ca8fbd4616501fa0
Message ID: <“swan.cl.cam.:219400:950824204042”@cl.cam.ac.uk>
Reply To: <199508242005.QAA03863@narn.atype.com>
UTC Datetime: 1995-08-24 20:41:18 UTC
Raw Date: Thu, 24 Aug 95 13:41:18 PDT

Raw message

From: Piete Brooks <Piete.Brooks@cl.cam.ac.uk>
Date: Thu, 24 Aug 95 13:41:18 PDT
To: jered@MIT.EDU
Subject: Re: SSL CHALLENGE: 0.11 borken?
In-Reply-To: <199508242005.QAA03863@narn.atype.com>
Message-ID: <"swan.cl.cam.:219400:950824204042"@cl.cam.ac.uk>
MIME-Version: 1.0
Content-Type: text/plain


>    I had several machines running brloop 0.03 and brclient 0.11, and
> none of them appeared to be sending ACKs back to the server.

The probability of getting an ACK back is indeed small :-((

> Is this a known problem?

yes -- this is why all the WWW pages and my messages to cypherpunks say
"please get the latest code" ...

> It might account for the large number of unACKed keyspaces.

I fear so :-((

> brloop 0.04 and brclient 0.12 seem to work well, however.

Whew !

brclient 0.13 is now out to try to track down the problem which is causing the
comgestion ...

It seems that old brloop's keep calling the server, and sending HELO, COMM and
then QUIT (Greet, Introduce, Part) i.e. not actually do anything :-(
[ Some only manage "HELO" or "HELO COMM" ]

These systems are hammering the server, and I suspect are the cause of the
congestion. I have added more logging to brclient 0.13, and it has caused the
problem to disappear ...





Thread