From owner-freebsd-current Fri Jan 8 07:55:36 1999 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA09854 for freebsd-current-outgoing; Fri, 8 Jan 1999 07:55:36 -0800 (PST) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from pau-amma.whistle.com (s205m64.whistle.com [207.76.205.64]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA09849 for ; Fri, 8 Jan 1999 07:55:35 -0800 (PST) (envelope-from dhw@whistle.com) Received: (from dhw@localhost) by pau-amma.whistle.com (8.9.1/8.9.1) id HAA07883 for freebsd-current@FreeBSD.ORG; Fri, 8 Jan 1999 07:54:06 -0800 (PST) (envelope-from dhw) Date: Fri, 8 Jan 1999 07:54:06 -0800 (PST) From: David Wolfskill Message-Id: <199901081554.HAA07883@pau-amma.whistle.com> Subject: Re: inetd problem Cc: freebsd-current@FreeBSD.ORG In-Reply-To: Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >From: Dag-Erling Smorgrav >Date: 08 Jan 1999 14:36:16 +0100 >No. People are constantly confused over this issue. There is (was) a >bug in inetd related to the way it handles signals, which caused it to >output "junk pointer: too low to make sense" (or sometimes "too high >to make sense") when under heavy load. It is not the same bug as the >dying daemons bug. There have been two attemps to fix the inetd bug: >one by Matt Dillon in rev. 1.42 and 1.43, and one by me (based on >patches submitted by Graham Wheeler) in rev. 1.44 and 1.45. I haven't >heard any complaints about the inetd bug lately, so I'll tentatively >postulate that I succeeded. One of my colleagues (Doug Ambrisko) has a desktop machine which has been exhibiting the symptom as recently as 1.43, thankss to the combination of things he does with the (abused?) machine and the load I put on it in the wee small hours of the morning using amanda (driving dump) to back it up. We were closed for the week between Christmas & New Year's (i.e., the first 8 days of Christmas), and he didn't get back to the office 'til the following Tuesday. I ran backups that week anyhow, but that wasn't enough to trigger the symptoms. The latest recurrence of the problem was yesterday (using the 1.43 rev. of the code); he said he would be applying DES' patches; I *thought* he meant that afternoon. Based on the results from this morning's backups, it looks as if he may have neglected to re-start inetd; I'll chat with him once he gets in. (Foggy here this AM, so he may be delayed getting in.) Historically, the failures have taken from 3-7 days to recur under the load mentioned above. david -- David Wolfskill UNIX System Administrator dhw@whistle.com voice: (650) 577-7158 pager: (650) 371-4621 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message