From owner-freebsd-hackers Sat Jan 1 22:55:14 2000 Delivered-To: freebsd-hackers@freebsd.org Received: from mass.cdrom.com (castles512.castles.com [208.214.165.76]) by hub.freebsd.org (Postfix) with ESMTP id AE9BC14E67; Sat, 1 Jan 2000 22:55:11 -0800 (PST) (envelope-from msmith@mass.cdrom.com) Received: from mass.cdrom.com (localhost [127.0.0.1]) by mass.cdrom.com (8.9.3/8.9.3) with ESMTP id XAA09576; Sat, 1 Jan 2000 23:00:23 -0800 (PST) (envelope-from msmith@mass.cdrom.com) Message-Id: <200001020700.XAA09576@mass.cdrom.com> X-Mailer: exmh version 2.1.1 10/15/1999 To: Greg Lehey Cc: Mike Smith , Tom , freebsd-hackers@FreeBSD.ORG Subject: Re: softupdates and debug.max_softdeps In-reply-to: Your message of "Sat, 01 Jan 2000 15:18:26 +1030." <20000101151826.L1528@freebie.lemis.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 01 Jan 2000 23:00:22 -0800 From: Mike Smith Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > On Thursday, 30 December 1999 at 22:34:04 -0800, Mike Smith wrote: > >> That is interesting. So I guess the conclusion to this is, softupdates > >> is useful for bursty IO, but not sustained because it can get far behind > >> until it eventually reaches the point where the machine reboots silently. > >> I guess the delay until reboot is dependent on the size of max_softdeps. > >> If it is big, it takes a while. > > > > I mentioned this a while back in the context of suspended I/O (in this > > case, a RAID array busy dealing with a failed disk). There wasn't much > > interest in dealing with it evinced at that point. > > On a related topic, I've taken to limiting the number of outstanding > transactions in Vinum, mainly to try to hunt down some strange > consistency problems when a very large number of transactions were > outstanding (for those of you who have been following this, this was > the "NULL b_biodone" syndrome). I still think there is a problem > hidden in the system which causes this. I've stopped seeing these in the Mylex driver after I stopped assuming that splbio() woult block re-entry into the strategy handler. In fact, if we had a lock-and-mask-interrupt primitive I wouldn't be using spl-anything, and I think that's the way we want to go. -- \\ Give a man a fish, and you feed him for a day. \\ Mike Smith \\ Tell him he should learn how to fish himself, \\ msmith@freebsd.org \\ and he'll hate you for a lifetime. \\ msmith@cdrom.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message