Hi Emil, We are only using the 82598 with the dom1 hosts. They are running Centos 2.6.18-348.12.1 kernel with an updated ixgbe driver v3.17.3 downloaded from the Intel site.
We have ixgbe blacklisted on dom0 which is running Centos6 2.6.32-358.18.1 I do see a lot of the messages like this in the libvirtd.log: 23210: error : virNetlinkEventCallback:339 : nl_recv returned with error: No buffer space available but I could not find any info on what that might mean, or how to correct. We are seeing the error count increase on the interface (ifconfig eth1 & ethtool -S) as the ping progresses, but no messages anywhere Thanks! --Larry On 09/03/2013 06:14 PM, Tantilov, Emil S wrote: >> -----Original Message----- >> From: laurence.schuler [mailto:[email protected]] >> Sent: Tuesday, September 03, 2013 1:36 PM >> To: [email protected] >> Subject: [E1000-devel] Fwd: libvirt-qemu/kvm passthru >> w/Intel 82598EB >> >> Hello, >> >> I am trying to configure two identical systems with the >> following setup. >> >> dom0: Centos 6 OS on Apple Xserve i7 server with PCI cards >> for 4GB FC (LSI FC949ES) and 10GB Ethernet (Intel 82598EB) >> >> dom1: Centos 5 OS with passthru mode of PCI 4GB FC and 10GB >> Ethernet (ixgbe and mptfc blacklisted in dom0) >> >> Everything starts ok, but networking in dom1 using the 10GB >> interface has extremely high rtt w/ping tests (20-30ms avg). > From the description above it seems you are using 82598 interfaces in both > dom0 and dom1, but you only see the issue on dom1? If that is the case, are > you only seeing the issue when dom1 runs Centos 5? What if you run dom1 with > Centos 6? > > I am not aware of any known issues with respect to passthru mode with qemu. > The driver should see and use the network adapter as usual and there should > be no special adjustments needed. We can try to reproduce this locally if you > could provide some additional information such as the distros and kernel > version of the host OS and the kernel versions in the VMs. > > Another thing to look at are any errors or warnings in dmesg when you bring > the VMs up. Also for the failing interface it is worth taking a look at the > stats while passing traffic via ethtool -S as this may give us an idea about > what is causing the performance drop. > > Thanks, > Emil > -- Laurence Schuler (Larry) [email protected] Systems Support ADNET Systems, Inc Scientific Visualization Studio http://svs.gsfc.nasa.gov NASA/Goddard Space Flight Center, Code 606.4 phone: 1-301-286-1799 Greenbelt, MD 20771 fax: 1-301-286-1634 Note: I am not a government employee and have no authority to obligate any federal, state or local government to perform any action or payment. ------------------------------------------------------------------------------ Learn the latest--Visual Studio 2012, SharePoint 2013, SQL 2012, more! Discover the easy way to master current and previous Microsoft technologies and advance your career. Get an incredible 1,500+ hours of step-by-step tutorial videos with LearnDevNow. Subscribe today and save! http://pubads.g.doubleclick.net/gampad/clk?id=58040911&iu=/4140/ostg.clktrk _______________________________________________ E1000-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/e1000-devel To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
