From owner-freebsd-stable@FreeBSD.ORG Fri Jul 11 03:37:51 2003 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id ED75737B401 for ; Fri, 11 Jul 2003 03:37:51 -0700 (PDT) Received: from mailhost.xciv.org (vantage.xciv.org [217.158.13.13]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2A34B43FB1 for ; Fri, 11 Jul 2003 03:37:51 -0700 (PDT) (envelope-from paul@xciv.org) Received: from localhost ([127.0.0.1] helo=xciv.org) by mailhost.xciv.org with esmtp id 19avHf-000LwP-00; Fri, 11 Jul 2003 11:37:47 +0100 To: freebsd-stable@freebsd.org Organization: XCIV, London UK Date: Fri, 11 Jul 2003 11:37:47 +0100 Message-ID: <84344.1057919867@xciv.org> From: Paul Civati Subject: SSH/v1 dead processes bug X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: paul@xciv.org List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 11 Jul 2003 10:37:52 -0000 Should I submit this as a proper bug, or will someone look at it? Basically, between 4.7-REL and 4.8-REL ssh was upgraded and now exhibits the following bug. If you connect with protocol v1 you get three ssh processes, one of which never dies, even after the ssh session has been closed. % ps aux | grep ssh root 219 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: paul [priv root 265 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: paul [priv root 580 0.0 0.3 2592 640 ?? Is Wed09PM 0:01.54 /usr/sbin/sshd root 584 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: root [priv root 592 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: root [priv root 1687 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: root [priv root 5077 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: root [priv root 40458 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: paul [priv root 40467 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: paul [priv root 41042 0.0 0.0 4076 0 ?? IW - 0:00.00 sshd: paul [priv root 79545 0.0 0.6 5292 1568 ?? I 10:33AM 0:00.03 sshd: paul [priv root 79547 0.0 0.6 4076 1532 ?? I 10:33AM 0:00.00 sshd: paul [priv paul 79548 0.0 0.7 5292 1656 ?? S 10:33AM 0:00.41 sshd: paul@ttyp0 root 80170 0.0 0.7 5292 1624 ?? I 11:07AM 0:00.03 sshd: paul [priv root 80172 0.0 0.6 4076 1592 ?? I 11:07AM 0:00.00 sshd: paul [priv paul 80173 0.0 0.8 5292 2044 ?? I 11:07AM 0:00.01 sshd: paul@ttypc root 80290 0.0 0.7 5292 1644 ?? I 11:12AM 0:00.02 sshd: paul [priv paul 80319 0.0 0.7 5292 1704 ?? I 11:14AM 0:00.01 sshd: paul@ttypd You'll see the dead ssh processes without any timestamp, two open v1 connections with thee processes, and one v2 connection with just two processes. The nasty problem with this I've found is that once you reach a certain number of these dead processes lying around, ssh seems to start refusing connections with: ssh_exchange_identification: Connection closed by remote host -Paul-