RE: Poor disk IO performance from Windows 2003 guest
There are a lot of variables here. Are you using virtio-blk devices and Windows guest drivers? No, those measurements were taken when I was using IDE emulated drives. To further clarify the NFS server is also software, not hardware, RAID-5. But your subtle hint and IBMs Best practices for KVM (http://publib.boulder.ibm.com/infocenter/lnxinfo/v3r0m0/topic/liaat/liaatbestpractices_pdf.pdf) were enough to galvanise me into action, so I have since reconfigured the guest to use Redhat Windows Virtio drivers instead. The results are much better, with 64MB writes on the system drive coming in at 39MB/s and reads 310MB/s. The second drive gives me 94MB/s for writes and 777MB/s for reads for a 64MB file. Again, that's wildy different results for two storage devices in the same guest, and it needs further investigation, but now the system is usable and I need to move on. For anyone reading this and looking for help to change the storage devices from IDE to virtio, the instructions I followed are at http://www.linux-kvm.com/content/redhat-54-windows-virtio-drivers-part-2-block-drivers Thanks for the hint! Kevin-- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Poor disk IO performance from Windows 2003 guest
On Wed, Mar 2, 2011 at 10:30 AM, Kevin Clark kevin.cl...@csoft.co.uk wrote: The results are much better, with 64MB writes on the system drive coming in at 39MB/s and reads 310MB/s. The second drive gives me 94MB/s for writes and 777MB/s for reads for a 64MB file. Again, that's wildy different results for two storage devices in the same guest, and it needs further investigation, but now the system is usable and I need to move on. Good to hear that you're seeing acceptable performance now. Stefan -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Poor disk IO performance from Windows 2003 guest
I'm getting wildly different disk IO performance from two storage devices in a Windows Server 2003 guest. I'm using diskgraf to test and get a max of 8MB/s for both reads and writes of a 64MB file on the system drive, but 101MB/s writes and 250MB/s reads on the second drive. The cache is enabled on both drives with driver name='qemu' cache='writeback'/. Prior to enabling the cache I was getting just 1.5MB/s reads/writes from the system drive. Both storage devices were created through virt-manager, and are raw image files access from an NFS server (SATA RAID5). Really hoping that someone can give me some pointers for fixing the problem. # modinfo kvm_intel filename: /lib/modules/2.6.18-194.32.1.el5/weak-updates/kmod-kvm/kvm-intel.ko license:GPL author: Qumranet version:kvm-83-164.el5_5.25 srcversion: 8C538112F6F8EC1E834351B depends:kvm vermagic: 2.6.18-194.17.1.el5 SMP mod_unload gcc-4.1 parm: bypass_guest_pf:bool parm: enable_vpid:bool parm: flexpriority_enabled:bool parm: enable_ept:bool parm: emulate_invalid_guest_state:bool # qemu-img info mercury_win2003_exchange2007_64.img image: mercury_win2003_exchange2007_64.img file format: raw virtual size: 29G (3145728 bytes) disk size: 29G # qemu-img info mercury_win2003_exchange2007_64_disk2.img image: mercury_win2003_exchange2007_64_disk2.img file format: raw virtual size: 20G (2097152 bytes) disk size: 20G # file mercury_win2003_exchange2007_64.img mercury_win2003_exchange2007_64.img: x86 boot sector, Microsoft Windows XP MBR, Serial 0x1ef01ee; partition 1: ID=0x7, active, starthead 1, startsector 63, 61416432 sectors # file mercury_win2003_exchange2007_64_disk2.img mercury_win2003_exchange2007_64_disk2.img: x86 boot sector, Microsoft Windows XP MBR, Serial 0x844035a3; partition 1: ID=0x7, starthead 1, startsector 63, 40949622 sectors, extended partition table (last)\011 Any thoughts/ideas? Thanks, Kevin -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Poor disk IO performance from Windows 2003 guest
On Tue, Mar 1, 2011 at 10:23 AM, Kevin Clark kevin.cl...@csoft.co.uk wrote: Any thoughts/ideas? There are a lot of variables here. Are you using virtio-blk devices and Windows guest drivers? Are you using hardware RAID5 on the NFS server? Could it be a network issue (contention during benchmark runs)? I'd start by benchmarking NFS on the host without running a virtual machine. Make sure you're getting acceptable performance and repeatable results there first. Stefan -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html