X-Original-To: alpine-devel@lists.alpinelinux.org Delivered-To: alpine-devel@lists.alpinelinux.org Received: from sp4.qcslink.com (sp4.qcslink.com [200.35.147.4]) by lists.alpinelinux.org (Postfix) with ESMTP id B29CD1FFFF0 for ; Mon, 18 Apr 2011 01:01:43 +0000 (UTC) Received: from sp4.qcslink.com (localhost.localdomain [127.0.0.1]) by sp4.qcslink.com (Postfix) with ESMTP id DBD091A46E for ; Sun, 17 Apr 2011 21:01:41 -0400 (EDT) Received: from [10.252.6.112] (c-67-188-63-60.hsd1.ca.comcast.net [67.188.63.60]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) (Authenticated sender: nangel@nothome.org) by sp4.qcslink.com (Postfix) with ESMTPSA id 8B2DB1A46D for ; Sun, 17 Apr 2011 21:01:41 -0400 (EDT) Message-ID: <4DAB8DFF.7060700@tetrasec.net> Date: Sun, 17 Apr 2011 18:03:59 -0700 From: Nathan Angelacos User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.2.15) Gecko/20110402 Icedove/3.1.9 X-Mailinglist: alpine-devel Precedence: list List-Id: Alpine Development List-Unsubscribe: List-Post: List-Help: List-Subscribe: MIME-Version: 1.0 To: Alpine-devel Subject: [alpine-devel] Alpinelinux openssh-hpn and vservers Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: ClamAV using ClamSMTP Just documenting something that I've noticed, don't know if this is the root cause or not. There are several guests running on a vserver host and each guest is running sshd Logging into a guest will /sometimes/ freeze up after the password or key exchange. That is, the server allocates the terminal and starts the command, but the client "locks up". The problem never happens on the vserver host. From the client side, ssh will work fine 80% of the time, but then won't work for several attempts, and then will work again. After upgrading to 17 Apr 2010 edge on the vserver and guests, the problem got worse. Noticed the following in the new openssh config file: > # the following are HPN related configuration options > # tcp receive buffer polling. disable in non autotuning kernels > # TcpRcvBufPoll yes > > # allow the use of the none cipher > #NoneEnabled no > > # disable hpn performance boosts. > # HPNDisabled no > > # buffer size for hpn to non-hpn connections > #HPNBufferSize 2048 After setting TcpRcvBufPoll to NO, the problem went away. --- Unsubscribe: alpine-devel+unsubscribe@lists.alpinelinux.org Help: alpine-devel+help@lists.alpinelinux.org ---