Image list

  • Ubuntu https://cloud-images.ubuntu.com/ .img是给QEUM和KVM用，.vhd是给Azure用，vmdk是给Vmware用。一般这些是在 OpenStack 这样的 IaaS 平台上面使用。这里描述如何在安装有 KVM 的普通 Linux 机器上面使用这些 Image。
  • Fedora https://cloud.fedoraproject.org/ Fedora cloud image，比 Ubuntu 兼容性好。
  • OpenSuse https://get.opensuse.org/leap/ Jeos is for cloud. Leap is normal version, Tumbleweed is rolling update. openSUSE MicroOS is a variant of openSUSE Tumbleweed and serves as a base of openSUSE Kubic, a Container as a Service platform. Guide of kubeadm’s installation container runtimes uses Tumbleweed. KVM and XEN version of JeOS image is cool: need interactive installation and it is fast. Doesn’t need extra disk for OS so boot image is enough. OpenStack-Cloud filesystem is xfs and 1G size. KVM and XEN filesystem is btrfs and 26G size. But Tumbleweed KVM version didn’t have igbvf driver. Crazy! Need zypper install kernel-default. It is 5.11.11. So what is stock or default kernel version?
  • Centos Stream https://cloud.centos.org/centos/
  • Rocky https://download.rockylinux.org/pub/rocky/8.4/images/
  • Linux Amazon Linux 2 https://aws.amazon.com/cn/amazon-linux-2/faqs/

Create VM

## Install a necessary packages 
$ sudo apt-get install kvm cloud-utils genisoimage 
## URL to most recent cloud image of 12.04 
$ img_url="http://cloud-images.ubuntu.com/server/releases/12.04/release" 
$ img_url="${img_url}/ubuntu-12.04-server-cloudimg-amd64-disk1.img" 
## download the image 
$ wget $img_url -O disk.img.dist 
## Create a file with some user-data in it 
$ cat > my-user-data <<EOF 
#cloud-config 
password: passw0rd 
chpasswd: { expire: False } 
ssh_pwauth: True
ssh_authorized_keys: 
  - ssh-rsa AAAAB3NzaC1yc2EAAAADABUB... fan@fandeiMac.lan
EOF 
## convert the compressed qcow file downloaded to a uncompressed qcow2 
$ qemu-img convert -O qcow2 download-disk.img disk.qcow2 
## create the disk with NoCloud data on it. 
$ cloud-localds my-seed.img my-user-data 
## if command not exist, run: genisoimage  -output seed.iso -volid cidata -joliet -rock user-data meta-data 
## Create a delta disk to keep our .orig file pristine 
$ qemu-img create -f qcow2 -b disk.qcow2 -F qcow2 disk-new.img 
## boot a kvm 
$ kvm -net nic -net user -hda disk.img -hdb my-seed.img -m 512

密码似乎不能用 123123 这种简单的。完整的 user-data spec 可以参考 http://cloudinit.readthedocs.io/en/latest/topics/examples.html，更多操作指南参考 KVM / libvirt。

对于 Centos Steam 和 Kubic，上面配置无法 login，这里 的 cloud-init.iso 可以用，原因未知。

上面配置要是能设置 static IP 和 hostname 就好了，有点怀念 Vagrant。

网络

详细配置看 http://cloudinit.readthedocs.io/en/latest/topics/network-config-format-v2.html https://gist.github.com/smoser/635897f845f7cb56c0a7ac3018a4f476 这个提供了样例和调试的方法。 其中match的方法对于我来说没啥用，因为mac地址是virtual function interface动态生成的，不固定。用match driver igbvf，发现： Kernel driver name, corresponding to the DRIVER udev property. Globs are supported. Matching on driver is only supported with networkd.

找了一圈，只有Fedora用的是systemd networkd，其他用的都是/etc/network/interface 这种老式的方式。 但是其cloud-init log里面有错误：AttributeError: ‘NoneType’ object has no attribute ‘iter_interfaces’ 嗯，centos呢？也是一样的错误。发现 centos7 (kernel v3.10) vm 启动很快，比 ubuntu 16.04 快多了。16.04 cloud image hangs at first boot，原来这个和 TTY有关系。但是我看 virt-manager 默认已经帮我生成好了啊。又尝试了下 Debian cloud image，也很快。 仔细查看上面 v2 文档，发现其需要 netplan，这样只有 Ubuntu 17.10 支持了？v1 我试过是可以的，但是其只能用 mac 地址匹配。

为什么每次启动 vm 后其virtual function mac地址会变呢？算了，最后我还是自己修改 os 配置为静态 ip 地址。对于 Centos，要删除 /etc/sysconfig/network-scripts/ifcfg-ens3 文件里面的 HWADDR=xxx 和 BOOTPROTO=dhcp。

Resize image

Ubuntu cloud image’s default size is 2GB. You can use qemu-img info a.img to check details. To resize it:

  1. Convert the qcow2 image to raw format qemu-img convert -O raw guest.img guest.raw
  2. Resize the raw file qemu-img resize guest.raw 10G
  3. Convert it back to qcow2(this will create 1G file, 1G = 10G * 0.10) qemu-img convert -O qcow2 -o compat=0.10 guest.raw guest.img.10g
  4. Power on vm and resize file system within the vm
    • run parted /dev/vda to edit the partition. set resize = “-1” (minus 1 means 1 sector from end of disk)
    • run resize2fs /dev/vda1

Using this way, ubuntu need lots of time to boot. It hang on “ systemd:started journal service”. I thought it was because I used virtIO disk or a small virtual image as disk. But at last it turn out Centos use xfs filesystem, after reboot, disk was enlarged.

Check snapshot file details:

fan@ubuntu-gg:~$ qemu-img info vm2.img
image: vm2.img
file format: qcow2
virtual size: 10G (10737418240 bytes)
disk size: 196K
cluster_size: 65536
backing file: ubuntu-16.04-server-cloudimg-amd64-disk1.img.10g
Format specific information:
    compat: 1.1
    lazy refcounts: false
    refcount bits: 16
    corrupt: false

It is small now but increase fast and soon it will be same size as original file.

Useful commands

virt-install --import --disk ubuntu1.img,bus=virtio --disk seed.img,device=cdrom --name vm1 --ram 8192 --nographics

https://quantum-integration.org/posts/install-cloud-guest-with-virt-install-and-cloud-init-configuration.html –nographics option forces virt-install to redirect the console output to the terminal window. After successful boot you get to the vm promt. And you can still access the text console by Virt Manager.

virsh console vm to connect running vm. The default escape key - to exit: ^[ ( Ctrl + ] )

Installing a KVM Guest OS from the Command-line (virt-install) command reference

KVM Install from Console

virt-install -n nwtest --description Test -r 512 --vcpus 1 \ 
   --location http://ftp.ca.debian.org/debian/dists/squeeze/main/installer-amd64/ \ 
   --os-type=linux --os-variant=debiansqueeze\ 
   --disk /srv/virtual/nwtest,device=disk,bus=virtio,size=3 \ 
   --network bridge=br0,model=virtio \ 
   --autostart \ 
   --nographics \ 
   -x "console=tty0 console=ttyS0,115200n8" 
Directly install from scratch, step by step. the location parameter specific kernel and initrd address. It is net-install indeed.

virsh domxml-to-native this command can convert virt instance into a vm creation command.