From owner-freebsd-current@FreeBSD.ORG Tue Apr 8 09:42:17 2003 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3DE2437B401 for ; Tue, 8 Apr 2003 09:42:17 -0700 (PDT) Received: from gs166.sp.cs.cmu.edu (GS166.SP.CS.CMU.EDU [128.2.205.169]) by mx1.FreeBSD.org (Postfix) with SMTP id 9DD4343FBF for ; Tue, 8 Apr 2003 09:42:16 -0700 (PDT) (envelope-from dpelleg@gs166.sp.cs.cmu.edu) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <16018.64453.390850.48895@gs166.sp.cs.cmu.edu> Date: Tue, 8 Apr 2003 12:41:41 -0400 To: Terry Lambert In-Reply-To: <3E92E509.A7A43986@mindspring.com> References: <3E92E509.A7A43986@mindspring.com> X-Mailer: VM 6.90 under 21.1 (patch 14) "Cuyahoga Valley" XEmacs Lucid From: Dan Pelleg Sender: dpelleg+@gs166.sp.cs.cmu.edu cc: current@freebsd.org Subject: Re: NIS exhausts system resources X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: Dan Pelleg List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 08 Apr 2003 16:42:17 -0000 Terry Lambert writes: > Dan Pelleg wrote: > > When does this happen, you ask? I triggered it this morning by booting the > > machine when the NIS server was down. I had also seen it in the past when > > configuring NIS, and it happened as soon as I set the domainname. Any > > ideas? I can provide packet captures on request, however note the failure > > where the server is down. > > Historical behaviour when the NIS server is down has been for the > client machines to hang until the NIS server is back up. > > FreeBSD doesn't serialize NIS requests trough a single local daemon, > so it doesn't hang "like it's supposed to". > > It's probably that you could reorder the source code to ensure that > no file opens (other than sockets) are held across an NIS request; > this would certainly reduce pressure on the number of open files. > > Note that since you are SSH'ing in, and have other processes running, > most likely, you just need to increase "MAXUSERS" to increase the > maximum number of open files, and let them hang simultaneously, > instead of running out of descriptors. > I'm not sure I understand this. First, if the server is down when the client boots, this will stop the client's boot sequence (or at least slow it significantly down). sshd isn't given a chance to start. The only sign of life from the machine at this point is ICMP replies and the sound of the grinding disk (probably syslogd). Next, this is a wimpy machine. 233Mhz processor and 64MB of RAM. I don't know if that matters or not, but I'm not confident of my ability to tweak the memory usage on it. > In general, NIS servers are not supposed to go down, ever. > > -- Terry