trent/trentdocs_website
1
0
Fork 0
mirror of https://github.com/TrentSPalmer/trentdocs_website.git synced 2025-11-03 22:28:25 -08:00
Code Issues Projects Releases Wiki Activity

extend stupid_kvm_tricks

Browse Source
This commit is contained in:
Trent Palmer
2018-02-11 08:33:05 -08:00
parent 632b56756d
commit 1721d26e54
5 changed files with 87 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
site/search/search_index.json
View File

@@ -582,7 +582,7 @@
},
{
"location": "/stupid_kvm_tricks/",
"text": "Stupid KVM Tricks\n\n\nvirt-install ubuntu16.04\n\n\nCreate the disk image\n\n\nqemu-img create -f qcow2 /var/lib/libvirt/images/xenial.qcow2 20G\n\n\nCommand to run the install\n\n\nvirt-install \\\n --name xenial \\\n --ram 4096 \\\n --disk path=/var/lib/libvirt/images/xenial.qcow2,size=20 \\\n --vcpus 4 \\\n --os-type linux \\\n --os-variant ubuntu16.04 \\\n --network bridge=br0 \\\n --graphics none \\\n --console pty,target_type=serial \\\n --location ./ubuntu-16.04.3-server-amd64.iso \\\n --extra-args 'console=ttyS0,115200n8 serial'\n\n\n\n\nChange the Network Interface\n\n\nbr0 gets addresses from the network router, but what if you want\nyour vm to have be on the virbr0 192.168.122.0/24 subnet?\n\n\nvirsh edit xenial\n\n\nAnd then 'J' all the way down to the bottom, change the interface name from br0 to\nvirbr0, \n\n\nvirsh start xenial\n\n\nand then look for the machine with nmap\n\n\nnmap -sn 192.168.122.0/24\n\n\nClone the VM\n\n\nIn this case we don't have to pre-allocate the disk image because virt-clone will do that\nfor us.\n\n\nvirt-clone --original xenial --name xenial-clone \\\n --file /var/lib/libvirt/images/xenial-clone.qcow2\n\n\n\n\nClone the VM to another Machine\n\n\nFirst dump the xml that defines the virtual machine.\n\n\nvirsh dumpxml xenial > xenial.xml\n\n\n\n\nThen copy both \nxenial.xml\n and \nxenial.qcow2\n to the new host machine. On the new kvm\nhost you'll want to at least make sure your vm has the correct CPU architecture.\nThe command to get a list of supported kvm cpu architectures is:\n\n\nvirsh cpu-models <arch>\n# i.e.\nvirsh cpu-models x86_64\n\n\n\n\nAfter you edit \nxenial.xml\n and update the correct cpu architecture, mv \nxenial.qcow2\n\nto \n/var/lib/libvirt/images/\n, clone it. \nvirt-clone\n will handle generating new\nmac addresses for the network interfaces.\n\n\n <cpu mode='custom' match='exact'>\n <model fallback='allow'>Haswell-noTSX</model>\n </cpu>\n# i.e. change to above to\n <cpu mode='custom' match='exact'>\n <model fallback='allow'>SandyBridge</model>\n </cpu>\n\n\n\n\n\nvirt-clone --original-xml xenial.xml --name xenial-clone \\\n --file /var/lib/libvirt/images/xenial-clone.qcow2\n\n\n\n\nWhat is the os-type and os-variant type names?\n\n\nosinfo-query os\n\n\nmisc\n\n\n\n\nStart the vm \nvirsh start xenial\n \n\n\nList all the vms \nvirsh list --all\n \n\n\nStop the vm \nvirsh destroy xenial\n \n\n\nDelete the vm \nvirsh undefine xenial\n \n\n\n\n\nvirsh help\n\n\nThe \nvirsh help\n command returns a long chart of help information. But each section has\na keyword.\n\n\nTake for instance the command \nvirsh help monitor\n. From this we\nsee the \ndomiflist\n subcommand (among others). Unfortunately \ndomifaddr\n doesn't seem to\nwork on the Ubuntu:16.04 host, but there are other ways to find the ip address of\na virtual machine.\n\n\nSo now if you want to see what host interface the vm \nxenial\n is attached to,\ntype. \n\n\nvirsh domiflist xenial\n\n\n\n\nwhich returns:\n\n\nInterface Type Source Model MAC\n-------------------------------------------------------\nvnet1 bridge virbr0 virtio 52:54:00:58:bf:75\n\n\n\n\nSo now we can find the address of virbr0 on the host machine.\n\n\nifconfig virbr0\n\n\n\n\nwhich returns:\n\n\nvirbr0 Link encap:Ethernet HWaddr 52:54:00:38:87:38 \n inet addr:192.168.122.1 Bcast:192.168.122.255 Mask:255.255.255.0\n UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1\n RX packets:1351 errors:0 dropped:0 overruns:0 frame:0\n TX packets:3037 errors:0 dropped:0 overruns:0 carrier:0\n collisions:0 txqueuelen:1000 \n RX bytes:232346 (232.3 KB) TX bytes:502916 (502.9 KB)\n\n\n\n\nand thus we know what subnet to scan with nmap to find the ip address of the vm\n\n\nnmap -sn 192.168.122.0/24",
"text": "Stupid KVM Tricks\n\n\nvirt-install ubuntu16.04\n\n\nCreate the disk image\n\n\nqemu-img create -f qcow2 /var/lib/libvirt/images/xenial.qcow2 20G\n\n\nCommand to run the install\n\n\nvirt-install \\\n --name xenial \\\n --ram 4096 \\\n --disk path=/var/lib/libvirt/images/xenial.qcow2,size=20 \\\n --vcpus 4 \\\n --os-type linux \\\n --os-variant ubuntu16.04 \\\n --network bridge=br0 \\\n --graphics none \\\n --console pty,target_type=serial \\\n --location ./ubuntu-16.04.3-server-amd64.iso \\\n --extra-args 'console=ttyS0,115200n8 serial'\n\n\n\n\nChange the Network Interface\n\n\nbr0 gets addresses from the network router, but what if you want\nyour vm to have be on the virbr0 192.168.122.0/24 subnet?\n\n\nvirsh edit xenial\n\n\nAnd then 'J' all the way down to the bottom, change the interface name from br0 to\nvirbr0, \n\n\nvirsh start xenial\n\n\nand then look for the machine with nmap\n\n\nnmap -sn 192.168.122.0/24\n\n\nClone the VM\n\n\nIn this case we don't have to pre-allocate the disk image because virt-clone will do that\nfor us.\n\n\nvirt-clone --original xenial --name xenial-clone \\\n --file /var/lib/libvirt/images/xenial-clone.qcow2\n\n\n\n\nClone the VM to another Machine\n\n\nFirst dump the xml that defines the virtual machine.\n\n\nvirsh dumpxml xenial > xenial.xml\n\n\n\n\nThen copy both \nxenial.xml\n and \nxenial.qcow2\n to the new host machine. On the new kvm\nhost you'll want to at least make sure your vm has the correct CPU architecture.\nThe command to get a list of supported kvm cpu architectures is:\n\n\nvirsh cpu-models <arch>\n# i.e.\nvirsh cpu-models x86_64\n\n\n\n\nAfter you edit \nxenial.xml\n and update the correct cpu architecture, mv \nxenial.qcow2\n\nto \n/var/lib/libvirt/images/\n, clone it. \nvirt-clone\n will handle generating new\nmac addresses for the network interfaces.\n\n\n <cpu mode='custom' match='exact'>\n <model fallback='allow'>Haswell-noTSX</model>\n </cpu>\n# i.e. change to above to\n <cpu mode='custom' match='exact'>\n <model fallback='allow'>SandyBridge</model>\n </cpu>\n\n\n\n\n\nvirt-clone --original-xml xenial.xml --name xenial-clone \\\n --file /var/lib/libvirt/images/xenial-clone.qcow2\n\n\n\n\nWhat is the os-type and os-variant type names?\n\n\nosinfo-query os\n\n\nmisc\n\n\n\n\nStart the vm \nvirsh start xenial\n \n\n\nList all the vms \nvirsh list --all\n \n\n\nStop the vm \nvirsh destroy xenial\n \n\n\nDelete the vm \nvirsh undefine xenial\n \n\n\n\n\nvirsh help\n\n\nThe \nvirsh help\n command returns a long chart of help information. But each section has\na keyword.\n\n\nTake for instance the command \nvirsh help monitor\n. From this we\nsee the \ndomiflist\n subcommand (among others). Unfortunately \ndomifaddr\n doesn't seem to\nwork on the Ubuntu:16.04 host, but there are other ways to find the ip address of\na virtual machine.\n\n\nSo now if you want to see what host interface the vm \nxenial\n is attached to,\ntype. \n\n\nvirsh domiflist xenial\n\n\n\n\nwhich returns:\n\n\nInterface Type Source Model MAC\n-------------------------------------------------------\nvnet1 bridge virbr0 virtio 52:54:00:58:bf:75\n\n\n\n\nSo now we can find the address of virbr0 on the host machine.\n\n\nifconfig virbr0\n\n\n\n\nwhich returns:\n\n\nvirbr0 Link encap:Ethernet HWaddr 52:54:00:38:87:38 \n inet addr:192.168.122.1 Bcast:192.168.122.255 Mask:255.255.255.0\n UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1\n RX packets:1351 errors:0 dropped:0 overruns:0 frame:0\n TX packets:3037 errors:0 dropped:0 overruns:0 carrier:0\n collisions:0 txqueuelen:1000 \n RX bytes:232346 (232.3 KB) TX bytes:502916 (502.9 KB)\n\n\n\n\nand thus we know what subnet to scan with nmap to find the ip address of the vm\n\n\nnmap -sn 192.168.122.0/24\n\n\n\n\nSnapshots\n\n\nCreate snapshot of vm \ndcing\n\n\nvirsh snapshot-create-as --domain dcing --name dcing-snap0\n\n\n\n\nBut you don't need to name your snapshots because they are listed by time.\n\n\nvirsh snapshot-create --domain dcing\n\n\n\n\nList snapshots for vm \ndcing\n\n\nvirsh snapshot-list --domain dcing\n\n Name Creation Time State\n------------------------------------------------------------\n 1518366561 2018-02-11 08:29:21 -0800 shutoff\n dcing-snap0 2018-02-11 08:22:57 -0800 shutoff\n\n\n\n\nRevert dcing to snap0\n\n\nvirsh snapshot-revert --domain dcing --snapshotname dcing-snap0\n\n\n\n\nDelete snapshot\n\n\nvirsh snapshot-delete --domain dcing --snapshotname dcing-snap0",
"title": "Stupid KVM Tricks"
},
{
@@ -624,6 +624,11 @@
"location": "/stupid_kvm_tricks/#virsh-help",
"text": "The virsh help command returns a long chart of help information. But each section has\na keyword. Take for instance the command virsh help monitor . From this we\nsee the domiflist subcommand (among others). Unfortunately domifaddr doesn't seem to\nwork on the Ubuntu:16.04 host, but there are other ways to find the ip address of\na virtual machine. So now if you want to see what host interface the vm xenial is attached to,\ntype. virsh domiflist xenial which returns: Interface Type Source Model MAC\n-------------------------------------------------------\nvnet1 bridge virbr0 virtio 52:54:00:58:bf:75 So now we can find the address of virbr0 on the host machine. ifconfig virbr0 which returns: virbr0 Link encap:Ethernet HWaddr 52:54:00:38:87:38 \n inet addr:192.168.122.1 Bcast:192.168.122.255 Mask:255.255.255.0\n UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1\n RX packets:1351 errors:0 dropped:0 overruns:0 frame:0\n TX packets:3037 errors:0 dropped:0 overruns:0 carrier:0\n collisions:0 txqueuelen:1000 \n RX bytes:232346 (232.3 KB) TX bytes:502916 (502.9 KB) and thus we know what subnet to scan with nmap to find the ip address of the vm nmap -sn 192.168.122.0/24",
"title": "virsh help"
},
{
"location": "/stupid_kvm_tricks/#snapshots",
"text": "Create snapshot of vm dcing virsh snapshot-create-as --domain dcing --name dcing-snap0 But you don't need to name your snapshots because they are listed by time. virsh snapshot-create --domain dcing List snapshots for vm dcing virsh snapshot-list --domain dcing\n\n Name Creation Time State\n------------------------------------------------------------\n 1518366561 2018-02-11 08:29:21 -0800 shutoff\n dcing-snap0 2018-02-11 08:22:57 -0800 shutoff Revert dcing to snap0 virsh snapshot-revert --domain dcing --snapshotname dcing-snap0 Delete snapshot virsh snapshot-delete --domain dcing --snapshotname dcing-snap0",
"title": "Snapshots"
}
]
}