virt

virt@lists.centos.org

1612 discussions

Re: [CentOS-virt] [CentOS] KVM and DHCP
by James B. Byrne 13 Mar '14

13 Mar '14

On Thu, March 13, 2014 15:49, Reindl Harald wrote: > > > Am 13.03.2014 20:43, schrieb James B. Byrne: >> CentOS-6.5 >> >> We have a KVM guest running MS-WinV7pro. This guest is joined to an Active >> Directory Domain. That domain provides DHCP to the members. The KVM guest >> does not obtain its IP from the domain but from the local host's qemu >> hypervisor instead. >> >> Is there anyway to get around this and have the guest MS-Win OS get its DHCP >> from the same place as the rest of the domain members? > > smells like you are using NAT for your guest instead > bridge it to the host network, the guest does not > know from where it gets DHCP infos and there should > not be more than one dhcpd on a network segment, anything > else leads to race-conditions - the faster dhcpd wins > > http://www.dedoimedo.com/computers/kvm-bridged.html > > maybe the article is outdated, try google as i did > https://www.google.at/search?q=kvm+nat+versus+bridging > > Yes, that was it. Apparently if one creates a kvm guest via the virt-manager gui application then NAT is automatically selected as the default. It may be that the other guests were originally hand-crafted via virsh and the br0 interface was specified for them, but I doubt that. I have been using virt-manager since day one on this project and I am getting a queasy feeling that virt-manager's behaviour has obscurely changed in significant ways since I first began using it. This is not the first problem that I have had creating new guests with recent versions of virt-manager which were not in evidence with earlier use. In any case, to fix this via virt-manager's __virtual hardware details__ interface one must select the __NIC__ and change the settings from <Virtual network default 'NAT'> to <Specify shared device name> and then input the bridge interface name <br0> in the text box that subsequently appears. One must also change the Device model: from <Hypervisor default> to <virtio>. Thanks for the pointer. -- *** E-Mail is NOT a SECURE channel *** James B. Byrne mailto:ByrneJB@Harte-Lyne.ca Harte & Lyne Limited http://www.harte-lyne.ca 9 Brockley Drive vox: +1 905 561 1241 Hamilton, Ontario fax: +1 905 561 0757 Canada L8E 3C3

1 0

KVM and DHCP
by James B. Byrne 13 Mar '14

13 Mar '14

CentOS-6.5 We have a KVM guest running MS-WinV7pro. This guest is joined to an Active Directory Domain. That domain provides DHCP to the members. The KVM guest does not obtain its IP from the domain but from the local host's qemu hypervisor instead. Is there anyway to get around this and have the guest MS-Win OS get its DHCP from the same place as the rest of the domain members? -- *** E-Mail is NOT a SECURE channel *** James B. Byrne mailto:ByrneJB@Harte-Lyne.ca Harte & Lyne Limited http://www.harte-lyne.ca 9 Brockley Drive vox: +1 905 561 1241 Hamilton, Ontario fax: +1 905 561 0757 Canada L8E 3C3

2 1

Are the "centos.org" AMIs in the EC2 marketplace "official"? locale breakage in 6.4 AMI
by Dan Pritts 13 Mar '14

13 Mar '14

Hi all, I've reviewed the mailing list archives and it looks like the "centos.org" AMIs in the marketplace are really official products of the centos project (whatever that means), but I wanted to ask explicitly to confirm. Presuming yes, I've run into an odd problem with the 6.4 x86_64 AMI (not updated). with a few packages, files are not being properly created in /usr/share/locale/xx/LC_XXXX when I install the RPM. I mostly don't care about the files themselves, but I worry about what else is wrong. I'll use subversion as my example below, but there are about half a dozen affected packages, including the coreutils package as shipped in the AMI (ami-bf5021d6). I have turned off selinux enforcement. RPM database shows the file: [root@dns-east svn]# rpm -ql subversion | grep de.LC /usr/share/locale/de/LC_MESSAGES/subversion.mo verify doesn't complain: [root@dns-east svn]# rpm --verify subversion [ just returns, no output ] RPM package contains the file: [root@dns-east svn]# rpm2cpio /tmp/subversion-1.6.11-10.el6_5.x86_64.rpm | cpio -t | grep de.LC_MESSAGES 23744 blocks ./usr/share/locale/de/LC_MESSAGES/subversion.mo [root@dns-east svn]# rpm2cpio /tmp/subversion-1.6.11-10.el6_5.x86_64.rpm | cpio --extract --make-directories 23744 blocks [root@dns-east svn]# ls -l ./usr/share/locale/de/LC_MESSAGES/ total 308 -rw-r--r--. 1 root root 312655 Mar 13 14:03 subversion.mo Uninstalling and reinstalling the RPM doesn't help. There are no crazy install scripts, either. [root@dns-east svn]# rpm -qp --scripts ../subversion-1.6.11-10.el6_5.x86_64.rpm postinstall scriptlet (using /bin/sh): # Register the snvserve service /sbin/chkconfig --add svnserve 2>&1 /sbin/ldconfig preuninstall scriptlet (using /bin/sh): if [ $1 = 0 ]; then /sbin/service svnserve stop > /dev/null 2>&1 || : /sbin/chkconfig --del svnserve > /dev/null 2>&1 || : fi postuninstall program: /sbin/ldconfig [root@dns-east svn]# -- Dan Pritts ICPSR Computing & Network Services University of Michigan +1 (734)615-7362

1 1

Remove Centos from AWS marketplace
by JacobV 09 Mar '14

09 Mar '14

https://forums.aws.amazon.com/thread.jspa?messageID=481859&#481859 https://forums.aws.amazon.com/thread.jspa?messageID=453572&#453572 This is a timebomb waiting to strike so many people who like do daily snapshot backups and keep them for few weeks and not realizing their snapshots are useless if they had accidentally mess up some boot related file earlier on. Another scenario you mess up the sudoers file or the root authorized keys - you'd have to loose a whole days data and go to previous nights restore just for single file error like that ? If AWS marketplace is unable to remove this hardcoded rule then it's only prudent to remove Centos from AWS marketplace and release it in the community section instead ? Thoughts ?

4 6

Xen4CentOS kernel panic on dom0 reboot
by Karl Johnson 05 Mar '14

05 Mar '14

Hello, I've been using Xen4CentOS for the last 3 months. It's working fine and dom0/domUs are stable but the server does a kernel panic when doing a reboot and the server has to be hard reset manually. It has kernel panic on the 3 last reboot. Please stand by while rebooting the system... INFO: task reboot:19800 blocked for more than 120 seconds "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message Call trace ... Picture of the call trace: http://i59.tinypic.com/169olt1.png CentOS: 6.5 Kernel: 3.10.23-11.el6.centos.alt.x86_64 Xen: 4.2.3 Server: Supermicro 6017R-WRF Any idea how to troubleshoot or fix this issue? We can't reboot the dom0 during the week because domUs (2) are in production and this server is in a datacenter. Kind regards, Karl

2 2

Xen4CentOS issue with running xen domains and a restart
by Johnny Hughes 25 Feb '14

25 Feb '14

There is an error in libvirt when rebooting or shutting down with a running xen domain on Xen4Centos6 When libvirtd tries to shutdown, it tries to suspend the running machine and fails with: VirDomainManagedSave error: this function is not supported by the connection driver Does anyone know how to patch libvirt 0.10.2.x to work around this issue.

1 0

disk io in guests causes soft lockups in guests and host processes
by Dennis Jacobfeuerborn 20 Feb '14

20 Feb '14

Hi, I have a strange phenomenon that I cannot readily explain so I wonder if anyone here can shed a light on this. The host system is a Dell r815 with 64 cores and 256G ram and has centos 6 installed. The five guests are also running centos 6 and are running as a hadoop cluster. The problem is that I see disk-io spikes in the vm's which then cause soft lockups in the guest but I also see hanging processes on the host as if the entire machine locks up for 30-60 seconds. Now I know that having all cluster members running on the same system isn't efficient and that I cannot expect good performance but what I was not expecting is that a guest make host processes hang. Does anyone have an idea what the issue could be here or how I can find out what cause for this behavior is? Regards, Dennis

3 2

Problem with cdrom device on guest
by James B. Byrne 13 Feb '14

13 Feb '14

I have a kvm guest running MS-WinV7 on a CentOS-6.5 host. The WinV7 guest was installed from a CD. It installed correctly and was updated and added to an MS Active Domain without difficulty. I was able to do work on it through the virt-manager console and I was able to shut the system down from the console without problem. I cannot get it to restart however. virsh start brws-ms-v7-37v.brockley.harte-lyne.ca error: Failed to start domain brws-ms-v7-37v.brockley.harte-lyne.ca error: cannot open file '/dev/sr0': No medium found I found this bugzilla report. https://bugzilla.redhat.com/show_bug.cgi?id=709585 Which is curiously marked as CLOSED DEFERRED. The last comment in that bug contains this: > Summary of findings so far: > > format=raw doesn't work with /dev/sr0. Omitting format=raw works. > format=host_device works. > > Apparently, virt-manager specifically asks for "raw". > > Things are different in RHEL-6: format=raw works, > format=host_device doesn't, format=host_cdrom does > (doesn't exist in RHEL-5). Now, in my configuration files there is no 'format' attribute found in any tag. I find 'type=raw' inside of many <device/> tags but it seems strange to me that a RedHat bug report would use such imprecise language so as to confuse the attribute 'type' with one named 'format' so I wonder if a syntax change has occurred and if that has introduced a regression? virsh edit brws-ms-v7-37v.brockley.harte-lyne.ca show this configuration entry for device sr0: 30 <disk type='block' device='cdrom'> 31 <driver name='qemu' type='raw'/> 32 <source dev='/dev/sr0'/> 33 <target dev='hdc' bus='ide'/> 34 <readonly/> 35 <address type='drive' controller='0' bus='1' target='0' unit='0'/> 36 </disk> Despite the inference I draw from the bug report, that format ~= type, one cannot in fact change the driver attribute 'type' to any value other than raw. Any attempt to do so results in this message: unsupported configuration: unknown driver format value 'host_cdrom' Failed. Try again? [y,n,f,?]: Further, the value 'f' is not permitted so one cannot forcibly save the configuration to determine if in fact it might work. Has anyone else run into this? What is the fix? -- *** E-Mail is NOT a SECURE channel *** James B. Byrne mailto:ByrneJB@Harte-Lyne.ca Harte & Lyne Limited http://www.harte-lyne.ca 9 Brockley Drive vox: +1 905 561 1241 Hamilton, Ontario fax: +1 905 561 0757 Canada L8E 3C3

1 0

documentation of snapshots
by Bill Campbell 06 Feb '14

06 Feb '14

I am looking for documentation on handling snapshots. The main tasks are: 1. Create a snapshot 2. Roll back a snapshot 3. Merge the snapshot into the base to start the cycle over. Currently I've done steps 1 and 2 into a single qcow2 image. I'm running a relatively old version of libvirtd on CentOS 5, libvirt-0.8.2-29.el5_9.1. Thanks. Bill -- INTERNET: bill(a)celestial.com Bill Campbell; Celestial Software LLC URL: http://www.celestial.com/ PO Box 820; 6641 E. Mercer Way Voice: (206) 236-1676 Mercer Island, WA 98040-0820 Fax: (206) 232-9186 Skype: jwccsllc (206) 855-5792 The question is, why are politicians so eager to be president? What is it about the job that makes it worth revealing, on national television, that you have the ethical standards of a slime-coated piece of industrial waste? -- Dave Barry, "On Presidential Politics"

1 0

Problem with lvm disks assigned to kvm guests
by C. L. Martinez 06 Feb '14

06 Feb '14

Hi all, I have a strange problem when I use lvm disks to expose to virtual guests (host is CentOS 6.5 x86_64). If I remove a kvm guest and all lvm disks attached to it, and I create a new kvm with another lvm disks that use the same disk space previously assigned to the previous kvm guest, this new guest sees all partitions and data. Creating new lvm volumes with different names to this new kvm doesn't resolves the problem. Any idea why?? Thanks.

6 10

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

virt