virt August 2016

virt@lists.centos.org

19 participants
12 discussions

No network using qemu-kvm-ev-2.3.0-31.el7.16.1
by C. L. Martinez 30 Aug '16

30 Aug '16

Hi all, To do some tests, I have installed a new CentOS7 host (fully patched) using qemu-kvm-ev-2.3.0-31.el7.16.1 instead of qemu-kvm's default package. And I have problems with network inside my guests. All guests can ping to CentOS host but nothing more: can't resolv dns names, can access via tcp to other hosts, etc ... Using same config on other CentOS7 host, but using default qemu-kvm package instead of qemu-kvm-ev-2.3.0-31.el7.16.1, all works ok. What am I doing wrong? Or is it a bug? Thanks -- Greetings, C. L. Martinez

3 4

Re: [CentOS-virt] No network using qemu-kvm-ev-2.3.0-31.el7.16.1
by C. L. Martinez 28 Aug '16

28 Aug '16

Ok, problem is not with qemu-kvm-ev-2.3.0-31.el7.16.1 package, it is with systemd-networkd. I have installed openvswitch and all works ok. Then, my question, is why is it not possible to use systemd-networkd for configure virtual switches to use them with KVM guests in CentOS 7 (using for other things, they seem work ok)?? In Ubuntu, Arch and Debian there is no problem with systemd-networkd configured bridges for KVM guests. Any idea why ?? Thanks. -- Greetings, C. L. Martinez

1 0

dracut-initqueue timeout with virt-install... but it works (kinda?)
by Matthew Crocker 25 Aug '16

25 Aug '16

Hello, I’m using virt-install to build a guest system with CentOS. The system boots up, times out with dracut-initqueue timeout and drops me into an emergency shell. If I exit the shell the install continues and I get a working machine. Any ideas? virt-install \ -n TEST \ -r 8192 \ --os-type=linux \ --disk=/vm-images/test.img,device=disk,bus=virtio,size=100,format=raw \ --ram=8192 \ --network bridge:br77 \ --location /home/CentOS-7-x86_64-Minimal-1511.iso \ --initrd-inject=/home/kickstart.cfg \ --extra-args="ks=file:/kickstart.cfg" \ --debug \ --accelerate kickstart.cfg: text install cdrom lang en_US.UTF-8 keyboard us network --onboot yes --device eth0 --bootproto dhcp --noipv6 rootpw testing123 firewall --service=ssh authconfig --enableshadow --passalgo=sha512 selinux --disable timezone --utc America/New_York bootloader --location=mbr zerombr clearpart --all --initlabel autopart reboot %packages --nobase @core %end — Matthew Crocker President - Crocker Communications, Inc. Managing Partner - Crocker Telecommunications, LLC E: matthew(a)corp.crocker.com E: matthew(a)crocker.com

2 1

Re: [CentOS-virt] how to enable auto-start on newer versions of Xen ?
by Craig Thompson 23 Aug '16

23 Aug '16

PJ, I didn't have virsh installed (libvirt-client). After having installed it, it still didn't see the process: virsh autostart testvm error: failed to connect to the hypervisor error: no valid connection error: Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory Then, after having installed libvirt, it gives essentially the same error. I was thinking that in iterations past, it was possible just to run the hypervisor, put a symlink in /etc/xen/auto, and VMs would be auto-started upon reboot. Maybe I'm wrong, and maybe those older boxes had libvirt installed on them, too. If I need to install libvirt, what does this mean overall? What am I changing? What other process needs to run or be started upon boot on a CentOS 7.x box? Thanks. -- Craig Thompson, President Caldwell Global Communications, Inc. -----Original message----- From: PJ Welsh <pjwelsh(a)gmail.com> Sent: Monday 15th August 2016 17:05 To: Discussion about the virtualization on CentOS <centos-virt(a)centos.org> Subject: Re: [CentOS-virt] how to enable auto-start on newer versions of Xen ? Try, "virsh autostart VMNAME" pjwelsh On Mon, Aug 15, 2016 at 2:54 PM, Craig Thompson <president(a)caldwellglobal.com <mailto:president@caldwellglobal.com> > wrote: Hello, In days past, all I had to do was create /etc/xen/auto and put a symlink in there to the config file for each VM I wanted to have started automatically. Since updating to 4.6, this doesn't work. Period. I'm having a hard time finding what needs to change in order to get VMs to auto start on reboot of a server. Can someone please point me to documentation that I have missed or share the gold nugget? -- Craig Thompson, President Caldwell Global Communications, Inc. _______________________________________________ CentOS-virt mailing list CentOS-virt(a)centos.org <mailto:CentOS-virt@centos.org> https://lists.centos.org/mailman/listinfo/centos-virt _______________________________________________ CentOS-virt mailing list CentOS-virt(a)centos.org https://lists.centos.org/mailman/listinfo/centos-virt

2 3

Nested KVM issue
by Laurentiu Soica 23 Aug '16

23 Aug '16

Hello, I have an OpenStack setup in virtual environment on CentOS 7. The baremetal has *nested KVM* enabled and 1 compute node as a VM. Inside the compute node I have multiple VMs running. After about every 3 days the VMs get inaccessible and the compute node reports high CPU usage. The qemu-kvm process for each VM inside the compute node reports full CPU usage. Please help me with some hints to debug this issue. Thanks, Laurentiu

3 22

Updating PXE documentation for CentOS wiki
by Nico Kadel-Garcia 20 Aug '16

20 Aug '16

I was just looking at the CentOS Wiki at https://wiki.centos.org/HowTos/PXE/PXE_Setup, and it's pretty seriously out of date. It neglects the existence of the "syslinux-tftpboot" RPM, and the hand editing of xinetd config files for tftpd ignores the availability of "chkconfig" and "service" to manage that service. For more sophisticated users, it doesn't mention "never, never, never set your default menu to install an OS by default, because you *will* accidentally wipe servers that select PXE boot first before local disk as their boot media". Nor does it mention the difficulties with PXE and NAT based virtual host, nor how to verify the TFTP service's basic operation, nor the difficulty of maintaining multiple PXE configs when the main rsync mirrors only publish the most recent CentOS. I'm happy to add that sort of thing to the wiki, but I'm finding the "Set up an account" page confusing. Is "FirstnameLastname" supposed to be your mandated login name?

2 2

qemu-kvm-ev-2.3.0-31.el7_2.21.1 available for testing on x86_64, ppc64le and aarch64
by Sandro Bonazzola 19 Aug '16

19 Aug '16

qemu-kvm-ev-2.3.0-31.el7_2.21.1 has been tagged for testing for CentOS Virt SIG and is already in testing repositories. It's now available for x86_64, ppc64le and aarch64. Please help testing and providing feedback, thanks. We plan to move to stable repo around Wednesday next week. ChangeLog since previous release: * Fri Aug 19 2016 Sandro Bonazzola <sbonazzo(a)redhat.com> - ev-2.3.0-31.el7_2.21 - Removing RH branding from package name * Tue Aug 02 2016 Miroslav Rezanina <mrezanin(a)redhat.com> - rhev-2.3.0-31.el7_2.21 - kvm-block-iscsi-avoid-potential-overflow-of-acb-task-cdb.patch [bz#1358997] - Resolves: bz#1358997 (CVE-2016-5126 qemu-kvm-rhev: Qemu: block: iscsi: buffer overflow in iscsi_aio_ioctl [rhel-7.2.z]) * Wed Jul 27 2016 Miroslav Rezanina <mrezanin(a)redhat.com> - rhev-2.3.0-31.el7_2.20 - kvm-virtio-error-out-if-guest-exceeds-virtqueue-size.patch [bz#1359731] - Resolves: bz#1359731 (EMBARGOED CVE-2016-5403 qemu-kvm-rhev: Qemu: virtio: unbounded memory allocation on host via guest leading to DoS [rhel-7.2.z]) * Wed Jul 20 2016 Miroslav Rezanina <mrezanin(a)redhat.com> - rhev-2.3.0-31.el7_2.19 - kvm-qemu-sockets-use-qapi_free_SocketAddress-in-cleanup.patch [bz#1354090] - kvm-tap-use-an-exit-notifier-to-call-down_script.patch [bz#1354090] - kvm-slirp-use-exit-notifier-for-slirp_smb_cleanup.patch [bz#1354090] - kvm-net-do-not-use-atexit-for-cleanup.patch [bz#1354090] - Resolves: bz#1354090 (Boot guest with vhostuser server mode, QEMU prompt 'Segmentation fault' after executing '(qemu)system_powerdown') * Fri Jul 08 2016 Miroslav Rezanina <mrezanin(a)redhat.com> - rhev-2.3.0-31.el7_2.18 - kvm-vhost-user-disable-chardev-handlers-on-close.patch [bz#1351892] - kvm-char-clean-up-remaining-chardevs-when-leaving.patch [bz#1351892] - kvm-sockets-add-helpers-for-creating-SocketAddress-from-.patch [bz#1351892] - kvm-socket-unlink-unix-socket-on-remove.patch [bz#1351892] - kvm-char-do-not-use-atexit-cleanup-handler.patch [bz#1351892] - Resolves: bz#1351892 (vhost-user: A socket file is not deleted after VM's port is detached.) * Tue Jun 28 2016 Miroslav Rezanina <mrezanin(a)redhat.com> - rhev-2.3.0-31.el7_2.17 - kvm-vhost-user-set-link-down-when-the-char-device-is-clo.patch [bz#1348593] - kvm-vhost-user-fix-use-after-free.patch [bz#1348593] - kvm-vhost-user-test-fix-up-rhel6-build.patch [bz#1348593] - kvm-vhost-user-test-fix-migration-overlap-test.patch [bz#1348593] - kvm-vhost-user-test-fix-chardriver-race.patch [bz#1348593] - kvm-vhost-user-test-use-unix-port-for-migration.patch [bz#1348593] - kvm-vhost-user-test-fix-crash-with-glib-2.36.patch [bz#1348593] - kvm-vhost-user-test-use-correct-ROM-to-speed-up-and-avoi.patch [bz#1348593] - kvm-tests-append-i386-tests.patch [bz#1348593] - kvm-vhost-user-add-ability-to-know-vhost-user-backend-di.patch [bz#1348593] - kvm-qemu-char-add-qemu_chr_disconnect-to-close-a-fd-acce.patch [bz#1348593] - kvm-vhost-user-disconnect-on-start-failure.patch [bz#1348593] - kvm-vhost-net-do-not-crash-if-backend-is-not-present.patch [bz#1348593] - kvm-vhost-net-save-restore-vhost-user-acked-features.patch [bz#1348593] - kvm-vhost-net-save-restore-vring-enable-state.patch [bz#1348593] - kvm-test-start-vhost-user-reconnect-test.patch [bz#1348593] - Resolves: bz#1348593 (No recovery after vhost-user process restart) -- Sandro Bonazzola Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com

1 0

how to enable auto-start on newer versions of Xen ?
by Craig Thompson 16 Aug '16

16 Aug '16

Hello, In days past, all I had to do was create /etc/xen/auto and put a symlink in there to the config file for each VM I wanted to have started automatically. Since updating to 4.6, this doesn't work. Period. I'm having a hard time finding what needs to change in order to get VMs to auto start on reboot of a server. Can someone please point me to documentation that I have missed or share the gold nugget? -- Craig Thompson, President Caldwell Global Communications, Inc.

4 3

Downgrading from Xen 4.6 to 4.5
by Francis Greaves 08 Aug '16

08 Aug '16

Dear All I have an up to date CentOS 7 with Xen 4.6 I have noticed in some posts here and there that PCI Passthrough is not working in 4.6 for some people, but is working in 4.5 How can I downgrade to Xen 4.5 so I can test this out? Many thanks Francis

3 3

Centos 7 newer kernel needed
by Laurentiu Soica 05 Aug '16

05 Aug '16

Hello, I am looking for a Linux kernel for Centos 7 that implements a feature introduced in kernel version 3.12: "nested EPT support to KVM's nested VMX." If anyone has used this feature with Centos 7, please let me know. Laurentiu

3 8

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

virt August 2016