From owner-freebsd-virtualization@freebsd.org Sun Oct 7 15:31:34 2018 Return-Path: Delivered-To: freebsd-virtualization@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id EFF1D10C4E5C for ; Sun, 7 Oct 2018 15:31:33 +0000 (UTC) (envelope-from ml@netfence.it) Received: from soth.netfence.it (net-2-44-121-52.cust.vodafonedsl.it [2.44.121.52]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (Client CN "mailserver.netfence.it", Issuer "Let's Encrypt Authority X3" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 6403B88AC1 for ; Sun, 7 Oct 2018 15:31:32 +0000 (UTC) (envelope-from ml@netfence.it) Received: from alamar.ventu (alamar.local.netfence.it [10.1.2.18]) (authenticated bits=0) by soth.netfence.it (8.15.2/8.15.2) with ESMTPSA id w97FVLXG074596 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NO) for ; Sun, 7 Oct 2018 17:31:27 +0200 (CEST) (envelope-from ml@netfence.it) X-Authentication-Warning: soth.netfence.it: Host alamar.local.netfence.it [10.1.2.18] claimed to be alamar.ventu From: Andrea Venturoli Subject: bhyve performance problem with SQL server To: freebsd-virtualization@freebsd.org Message-ID: Date: Sun, 7 Oct 2018 17:31:17 +0200 User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:60.0) Gecko/20100101 Thunderbird/60.0.1 MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit X-BeenThere: freebsd-virtualization@freebsd.org X-Mailman-Version: 2.1.27 Precedence: list List-Id: "Discussion of various virtualization techniques FreeBSD supports." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 07 Oct 2018 15:31:34 -0000 Hello. I've got a FreeBSD 11.2 box running vm-bhyve with two Windows virtual machines, each with an instance of MS SQL server. Everything works fine normally, but there's a specific operation which will impose an high CPU load on the machine and last several minutes, when it's expected to get the results in a few seconds. I think I have to pass this trouble to the "programming" people, but before I do that, I'd like to double-check this is not a system problem. Some notes: _ that specific operation will put the two VMs in touch with each other, as one will contact the DBMS on the other and import data; however, that's not the only procedure which does such a thing and the others are working fine; _ during the execution of such procedure, the load on the host rises, as both VMs show high CPU usage (in Task Manager), top shows both bhyve processes in kqread (but this happens all the time), disk activity doesn't seem so high; _ the box has a Quad-Core CPU and I assigned two cores to each VM (I have hw.vmm.topology.cores_per_package=4 in /boot/loader.conf); _ both VMs use virtio-net, with both tap interfaces connected to the same "virtual switch"; _ both VMs use a zvol for their HD. Before I sentence this is an application level problem, is there some improvment I could do to the above? Any check or test I should run? bye & Thanks av.