From owner-freebsd-questions@FreeBSD.ORG Mon Mar 17 21:08:47 2008 Return-Path: Delivered-To: questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id E8880106564A for ; Mon, 17 Mar 2008 21:08:47 +0000 (UTC) (envelope-from danger@FreeBSD.org) Received: from virtual.micronet.sk (smtp.micronet.sk [84.16.32.237]) by mx1.freebsd.org (Postfix) with ESMTP id 9FEBF8FC13 for ; Mon, 17 Mar 2008 21:08:47 +0000 (UTC) (envelope-from danger@FreeBSD.org) Received: from localhost (localhost [127.0.0.1]) by virtual.micronet.sk (Postfix) with ESMTP id 4A4C210E6AB for ; Mon, 17 Mar 2008 21:44:02 +0100 (CET) X-Virus-Scanned: by amavisd-new at virtual.micronet.sk Received: from virtual.micronet.sk ([127.0.0.1]) by localhost (virtual.micronet.sk [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sndQHh2AAusu for ; Mon, 17 Mar 2008 21:43:56 +0100 (CET) Received: from DANGER-PC (danger.mcrn.sk [84.16.37.254]) by virtual.micronet.sk (Postfix) with ESMTP id 76A6D10E6A4 for ; Mon, 17 Mar 2008 21:43:56 +0100 (CET) Date: Mon, 17 Mar 2008 21:45:11 +0100 From: Daniel Gerzo X-Mailer: The Bat! (v3.99.3) Professional Organization: The FreeBSD Project X-Priority: 3 (Normal) Message-ID: <14814001.20080317214511@rulez.sk> To: questions@freebsd.org MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Subject: interrupt storms X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: Daniel Gerzo List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 17 Mar 2008 21:08:48 -0000 Hello, recently, we have inserted a new NIC into our machine, and after a few hours of operation, the machine starts to suffer from interrupt storms. Mar 17 21:05:24 ha-web1 kernel: interrupt storm detected on "irq22:"; throttling interrupt source Mar 17 21:05:54 ha-web1 last message repeated 30 times Mar 17 21:07:54 ha-web1 last message repeated 120 times Mar 17 21:17:54 ha-web1 last message repeated 600 times Mar 17 21:27:54 ha-web1 last message repeated 600 times Here is the output of vmstat -i: interrupt total rate irq1: atkbd0 5 0 irq9: acpi0 1 0 irq16: ohci0 1 0 irq17: ohci1 ohci3 1 0 irq18: ohci2 ohci4 1 0 irq20: em0 22617825 1714 irq22: em1 atapci0 699549409 53036 cpu0: timer 26380374 2000 cpu1: timer 26380325 2000 Total 774927942 58751 Any idea, why em1 and atapci0 share the same irq? I don't think this is correct. Is there some way how to work this around? Thanks. -- Best regards, Daniel mailto:danger@FreeBSD.org