|
|
| Line 1: |
Line 1: |
| [[Category:Virtualization]] | | [[Category:Virtualization]] <!-- do not copy into FWN issue --> |
|
| |
|
| {{Anchor|Virtualization}} | | {{Anchor|Virtualization}} |
|
| |
|
| == Virtualization == | | == Virtualization == |
| In this section, we cover discussion on the @et-mgmnt-tools-list, @fedora-xen-list, and @libvirt-list of Fedora virtualization technologies. | | In this section, we cover discussion on the @et-mgmnt-tools-list, @fedora-xen-list, @libvirt-list and @ovirt-devel-list of Fedora virtualization technologies. |
|
| |
|
| Contributing Writer: [[User:Dale | Dale Bewley]] | | Contributing Writer: [[User:Dale | Dale Bewley]] |
| Line 12: |
Line 12: |
| [http://www.redhat.com/mailman/listinfo/et-mgmt-tools et-mgmt-tools list] | | [http://www.redhat.com/mailman/listinfo/et-mgmt-tools et-mgmt-tools list] |
|
| |
|
| ==== virt-manager and QEmu Disk Polling Logs ==== | | ==== ==== |
| Radek Hladik
| |
| noticed<ref>http://www.redhat.com/archives/et-mgmt-tools/2009-February/msg00019.html</ref>
| |
| "when {{package|virt-manager}} is running and polling VMs stats {{package|libvirt}} log in <code>/var/log/libivrt/qemu/vmname</code> is filling with messages" on the number of disk operations. After a day the log had grown to 100MB.
| |
| | |
| Daniel P. Berrange accepted<ref>http://www.redhat.com/archives/et-mgmt-tools/2009-February/msg00021.html</ref> on behalf of <code>libvirt</code>
| |
| and Cole Robinson
| |
| described<ref>http://www.redhat.com/archives/et-mgmt-tools/2009-February/msg00020.html</ref>
| |
| how to turn of disk polling in <code>virt-manager</code>.
| |
| | |
| <references />
| |
| | |
| ==== virt-viewer Persistance Through Guest Reboots ====
| |
| Daniel P. Berrange
| |
| said<ref>http://www.redhat.com/archives/et-mgmt-tools/2009-February/msg00004.html</ref> that by Fedora 11 {{package|virt-viewer}} will persist and wait for a guest to resume rather than exit when a guest reboots.
| |
| | |
| <references />
| |
| | |
| ==== virt-install Wait Indefinitely for Windows Guests ====
| |
| Since Windows reboots during installation, John Levon patched<ref>http://www.redhat.com/archives/et-mgmt-tools/2009-February/msg00005.html</ref> <code>virt-install</code> to wait 120 minutes while installing a Windows guest. After some discussion it was decided it should wait indefinitely instead.
| |
| | |
| <references /> | | <references /> |
|
| |
|
| Line 39: |
Line 19: |
| [http://www.redhat.com/mailman/listinfo/fedora-virt fedora-virt list]. | | [http://www.redhat.com/mailman/listinfo/fedora-virt fedora-virt list]. |
|
| |
|
| ==== Fedora Virt Status Update ==== | | ==== ==== |
| Mark McLoughlin composed<ref>http://www.redhat.com/archives/fedora-virt/2009-February/msg00061.html</ref> another informative weekly update on the status of virtualization development in Fedora. Some highlights included:
| |
| | |
| * A <code>pvmmu</code> problem casues some guest installs on an F11 Alpha host to <code>oops</code> during heavy network activity ({{bz|480822}})
| |
| * Work has begun on Fedora 11 virtualization [[Documentation_Virtualization_Beat|release notes]].
| |
| * The 0.6.0 release of {{package|libvirt}} was not completely without problems.
| |
| * All Fedora 11 virtualization features can be found [[:Category:F11_Virt_Features|all together]].
| |
| * The <code>KVM</code>/<code>QEmu</code> merge project has settled on a naming scheme.
| |
| * <code>KVM</code> PCI device assignment continues to have issues. The "core of the problem is that devices must be reset before being assigned if they have been previously used in the host."
| |
| * The addition of <code>bzImage</code> loading support to the <code>Xen</code> hypervisor is enabling users to build test Dom0 kernels.
| |
| * A detailed of accounting reveals the bug count going from 191 to 192.
| |
| | |
| <references />
| |
| | |
| ==== Merging KVM and QEmu Packages ==== | |
| A feature<ref>[[Features/KVM_and_QEMU_merge]]</ref> in the works for Fedora 11 is a merge of the {{package|kvm}} package with the {{package|qemu}} package. Glauber Costa recently took the first step in this process by creating a test
| |
| build<ref>http://koji.fedoraproject.org/koji/taskinfo?taskID=1105051</ref> and starting a very long thread<ref>http://www.redhat.com/archives/fedora-virt/2009-February/msg00000.html</ref> on naming of all the subpackages which will soon make up <code>QEmu</code>.
| |
| | |
| <references /> | | <references /> |
|
| |
|
| Line 63: |
Line 26: |
| [http://www.redhat.com/mailman/listinfo/fedora-xen fedora-xen list]. | | [http://www.redhat.com/mailman/listinfo/fedora-xen fedora-xen list]. |
|
| |
|
| ==== bzImage Dom0 Support in Rawhide Xen ==== | | ==== ==== |
| Pasi Kärkkäinen
| |
| announced<ref>http://www.redhat.com/archives/fedora-xen/2009-February/msg00001.html</ref>
| |
| the lastest {{package|xen}} builds in [[Releases/Rawhide|Rawhide]] support
| |
| <code>bzImage</code> compressed dom0 kernels. Xen previously only supported
| |
| <code>zImage</code> compressed kernels. This development was one of the preqequisite
| |
| work items for the [[Features/XenPvopsDom0|Xen pvops Dom0]] feature.
| |
| | |
| This good news was tempered by the fact that there is still no dom0 capable
| |
| {{package|kernel}} in Rawhide. However, such a <code>kernel</code> can be
| |
| built<ref>http://wiki.xensource.com/xenwiki/XenParavirtOps</ref> for testing.
| |
| Gerd Hoffmann
| |
| reports<ref>http://www.redhat.com/archives/fedora-virt/2009-February/msg00027.html</ref>
| |
| success doing just that. Such kernels are not yet stable enough for
| |
| use<ref>http://www.redhat.com/archives/fedora-virt/2009-February/msg00055.html</ref>.
| |
| | |
| <references />
| |
| | |
| ==== Test Dom0 Xen Kernel RPM Available ====
| |
| M A Young
| |
| built<ref>http://www.redhat.com/archives/fedora-xen/2009-February/msg00014.html</ref>
| |
| a dom0 capable <code>kernel</code> RPM. It's suitable only for testing; "use it very much at your own risk".
| |
| | |
| <references /> | | <references /> |
|
| |
|
| Line 92: |
Line 33: |
| [http://www.redhat.com/mailman/listinfo/libvir-list libvir-list]. | | [http://www.redhat.com/mailman/listinfo/libvir-list libvir-list]. |
|
| |
|
| ==== Fix for Fallout From Failed QEmu Guest Starts ==== | | ==== ==== |
| Daniel P. Berrange
| |
| fixed<ref>http://www.redhat.com/archives/libvir-list/2009-February/msg00104.html</ref> a series of events which manifested when a QEmu guest failed to start. Subsequent client connections would fail, CPU would rise to 100%, and <code>virsh</code> would hang. ({{bz|484414}})
| |
| | |
| <references />
| |
| | |
| ==== sVirt Patches to Merge in libvirt ====
| |
| Daniel J Walsh
| |
| looked<ref>http://www.redhat.com/archives/libvir-list/2009-February/msg00144.html</ref> at James Morris <code>sVirt</code><ref>http://fedoraproject.org/wiki/Features/SVirt_Mandatory_Access_Control</ref> patches for {{package|libvirt}}.
| |
| "James patch, allows <code>libvirt</code> to read the <code>SELinux</code> context out of the xml
| |
| database and execute {{package|qemu}} with the context. The second componant [sic] is to
| |
| pass the context of the image(s) and allow <code>libvirt</code> to not only set the
| |
| image, but also update the default labels on disk, so a relabel will not
| |
| change the context." Daniel J Walsh started working on this second component
| |
| and wondered if they were acceptable for committing to <code>libvirt</code> yet.
| |
| | |
| Daniel P. Berrange expressed<ref>http://www.redhat.com/archives/libvir-list/2009-February/msg00145.html</ref> satisfaction with how the patches integrate with <code>libvirt</code> adding
| |
| "If yourself & James are happy with what they're doing from a SELinux /
| |
| security model point of view, then there's no reason they shouldn't
| |
| be posted for final merge now."
| |
| | |
| <references />
| |
| | |
| ==== Manage iptables with libvirt ====
| |
| Karl Wirth asked<ref>http://www.redhat.com/archives/libvir-list/2009-February/msg00147.html</ref> "What if we could flexibly change the iptables rules for the different guests as they are deployed onto the node/host".
| |
| This thought was not new.<ref>http://fedoraproject.org/wiki/FWN/Issue138#Libvirt_and_Persistent_Iptables_Rules</ref>
| |
| | |
| David Lutterkort pointed<ref>http://www.redhat.com/archives/libvir-list/2009-February/msg00152.html</ref> out some of the thorny problems with this proposal including the fact that
| |
| "network devices may be directly assigned to guests - in that
| |
| case, we won't even see any of the packets the guest sends or receives".
| |
| Summarizing that "iptables management belongs into a higher-level management
| |
| app, like <code>ovirt</code><ref>http://ovirt.org/</ref>, not <code>libvirt</code>."
| |
| | |
| <references /> | | <references /> |
