From owner-freebsd-questions@FreeBSD.ORG Tue Jun 3 07:44:48 2008 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id EF22F106567C for ; Tue, 3 Jun 2008 07:44:48 +0000 (UTC) (envelope-from the-ghost@inbox.ru) Received: from mx44.mail.ru (mx44.mail.ru [195.239.211.10]) by mx1.freebsd.org (Postfix) with ESMTP id 8F8AC8FC16 for ; Tue, 3 Jun 2008 07:44:48 +0000 (UTC) (envelope-from the-ghost@inbox.ru) Received: from f107.mail.ru (f107.mail.ru [194.67.57.67]) by mx44.mail.ru (mPOP.Fallback_MX) with ESMTP id C5898380031B5 for ; Tue, 3 Jun 2008 11:31:12 +0400 (MSD) Received: from mail by f107.mail.ru with local id 1K3Qz3-000OTE-00 for freebsd-questions@freebsd.org; Tue, 03 Jun 2008 11:31:05 +0400 Received: from [85.140.79.205] by win.mail.ru with HTTP; Tue, 03 Jun 2008 11:31:05 +0400 From: The Ghost To: freebsd-questions@freebsd.org Mime-Version: 1.0 X-Mailer: mPOP Web-Mail 2.19 X-Originating-IP: [85.140.79.205] Date: Tue, 03 Jun 2008 11:31:05 +0400 Content-Type: text/plain; charset=koi8-r Content-Transfer-Encoding: 8bit Message-Id: X-Spam: Not detected X-Mras: OK Subject: very strange reaction of the md disks X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: The Ghost List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Jun 2008 07:44:49 -0000 Hello, Maybe I just don't understand something, but anyway, it looks really odd. I have a machine with 2 Gb RAM with a clean newly-installed FREEBSD-7.0-RELEASE. I make two 300-MB mdconfig disks and populate them, and when the first one was already full and the second one is only half-full, I get a kernel panic "not enough memory". I wonder where did 1,5 Gb go?? Well, I tried to reboot and allocate them with -o reserve to prevent the kernel panic, but the second one said that enough memory can't be allocated (though top still showed 1,5 Gb left). Well, I rebooted and tried to allocate one 1-Gb disk, but it also failed. Then I gave up on this and came to my other work, but when I loaded some small kernel module, I got a kernel panic!! What, didn't the disk allocation have to kinda free up the used memory when it fails?! And then the last thing. I reboot and copy a 220-m file from a samba-mounted directory to my hard drive (UFS2 with soft-updates). I was very surprised to see that one gigabyte of my memory suddenly became used up. I wonder if the soft-updat es really don't care about how much memory they use... A friend of mine told me that such a panic is really a thing to write a bugreport on it, but I know about this behaviour for few years already and so I guess I just don't know something about the memory allocation system, and, kinda, it really works just like that. Should I read something to understand why it's just like that or really write a report on some of these things?.. The Ghost