From owner-freebsd-questions Tue Jun 16 06:03:04 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id GAA01962 for freebsd-questions-outgoing; Tue, 16 Jun 1998 06:03:04 -0700 (PDT) (envelope-from owner-freebsd-questions@FreeBSD.ORG) Received: from elvis.vnet.net (elvis.vnet.net [166.82.1.5]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id GAA01944 for ; Tue, 16 Jun 1998 06:02:59 -0700 (PDT) (envelope-from rivers@dignus.com) Received: from dignus.com (ponds.vnet.net [166.82.177.48]) by elvis.vnet.net (8.8.8/8.8.4) with ESMTP id JAA12213; Tue, 16 Jun 1998 09:02:32 -0400 (EDT) Received: from lakes.dignus.com (lakes [10.0.0.3]) by dignus.com (8.8.5/8.8.5) with ESMTP id IAA03000; Tue, 16 Jun 1998 08:55:36 -0400 (EDT) Received: (from rivers@localhost) by lakes.dignus.com (8.8.7/8.6.9) id IAA02747; Tue, 16 Jun 1998 08:29:47 -0400 (EDT) Date: Tue, 16 Jun 1998 08:29:47 -0400 (EDT) From: Thomas David Rivers Message-Id: <199806161229.IAA02747@lakes.dignus.com> To: dg@root.com, freebsd-questions@FreeBSD.ORG, hua@nease.net, rivers@dignus.com, sysadmin@mfn.org Subject: RE: Is there any "race condition" in FFS? Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Well - let's see how I can allay those fears. First - the phenomenon appears to happen to a select few around the globe (I believe you would be the fourth or the fifth.) So, for the hundreds of thousands of FreeBSD users - this is not a problem. Second - although it was bothersome - I never actually lost any data. Third - it seems to happen during a protractedly large set of I/O operations (i.e. aging a news spool is the more popular way of getting this to happen.) Fourth - I've copied David Greenman on this. [David - this appears to be similar to the "dup alloc" problems.] He'll have even more good news. Fifth - there are several large ISPs and web sites happily running FreeBSD (who don't get this problem) with thousands of users. So, it's quite doable. I believe we can work through this. Can you provide more detail; such as: 1) The exact panic message, and what appears to be going on when you get it. 2) The panic traceback (you'll need to enable kernel dumping on reboot to get the kernel and core file.) 3) The version of FreeBSD you're using. > > OK, now you've got me nervous! Here I am about to begin turning mail > services here over to FBSD and Popper et al. (from NT with IMS - a really > great mail server whose mail flaw is running on NT ;-) and you send me > word of th "David Rivers Daily Memorial Panic"??? > > Is this phenomenon associated with any particular set of conditions > which I can avoid? This is *very* serious from my perspective: let's > face it Email is the *one* thing that will fire up an otherwise brain-dead > user to action! I certainly don't need 5000 daily complaints from angry > users and frustrated bosses :) > > Should I be sticking with our NT mail servers for now? In my experience, NT mail servers are not adequate to handle more than 10 people. Ours are constantly crashing; and when they do stay up, they seem to loose mail... I would, personally, prefer anything over NT for mail servers. > > Thanks, > J.A. Terranson > sysadmin@mfn.org > > > > > > Recently,some of our Server always down due to the FFS error. > > the situation can be describe below: > > Al - > > I believe you'll find similar situations in the FreeBSD mail > archives. Look for "daily panic", and the "Dave Rivers memorial > panic" in the freebsd-hackers mail archive. > > But, there is some good news. People have reported that the problem > appears to have gone away in version 3.0 (likely as a good side-effect > of other changes.) You might want to try using 3.0 on your server. > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message