Date: Fri, 2 May 2014 01:27:46 GMT From: Allan Jude <freebsd@allanjude.com> To: freebsd-gnats-submit@FreeBSD.org Subject: docs/189216: [patch] add a handbook section on hosting VMs with bhyve Message-ID: <201405020127.s421Rk5L039399@cgiserv.freebsd.org> Resent-Message-ID: <201405020130.s421U0sP031044@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
>Number: 189216
>Category: docs
>Synopsis: [patch] add a handbook section on hosting VMs with bhyve
>Confidential: no
>Severity: non-critical
>Priority: low
>Responsible: freebsd-doc
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: change-request
>Submitter-Id: current-users
>Arrival-Date: Fri May 02 01:30:00 UTC 2014
>Closed-Date:
>Last-Modified:
>Originator: Allan Jude
>Release: 10.0-STABLE
>Organization:
ScaleEngine Inc.
>Environment:
FreeBSD Trooper.HML3.ScaleEngine.net 10.0-STABLE FreeBSD 10.0-STABLE #0 Sat Mar 22 13:15:35 EDT 2014 root@Trooper.HML3.ScaleEngine.net:/usr/obj/media/10stable/sys/GENERIC amd64
>Description:
This patch adds a bhyve section to the virtualization chapter, and separates the virtualbox chapter.
Sponsored by: ScaleEngine Inc.
>How-To-Repeat:
>Fix:
Patch attached with submission follows:
Index: handbook/virtualization/chapter.xml
===================================================================
--- handbook/virtualization/chapter.xml (revision 44736)
+++ handbook/virtualization/chapter.xml (working copy)
@@ -20,6 +20,16 @@
<contrib>Contributed by </contrib>
</author>
</authorgroup>
+
+ <authorgroup>
+ <author>
+ <personname>
+ <firstname>Allan</firstname>
+ <surname>Jude</surname>
+ </personname>
+ <contrib>bhyve section by </contrib>
+ </author>
+ </authorgroup>
</info>
<sect1 xml:id="virtualization-synopsis">
@@ -1110,8 +1120,8 @@
</sect2>
</sect1>
- <sect1 xml:id="virtualization-host">
- <title>&os; as a Host</title>
+ <sect1 xml:id="virtualization-host-virtualbox">
+ <title>&os; as a Host with <application>VirtualBox</application></title>
<para><application>&virtualbox;</application> is an actively
developed, complete virtualization package, that is available
@@ -1273,7 +1283,310 @@
<screen>&prompt.root; <userinput>service devfs restart</userinput></screen>
</sect2>
+ </sect1>
+ <sect1 xml:id="virtualization-host-bhyve">
+ <title>&os; as a Host with
+ <application>bhyve</application></title>
+
+ <para>Starting with &os; 10.0-RELEASE the BSD licensed hypervisor
+ <application>bhyve</application> is part of the base system.
+ <application>bhyve</application> supports a number of guests
+ including &os;, OpenBSD, and many flavors of &linux;.
+ <application>bhyve</application> currently only supports a
+ serial console and does not emulate a graphical console.
+ <application>bhyve</application> is a legacy-free hypervisor,
+ meaning that instead of translating instructions, and manually
+ managing memory mappings, it relies on the virtualization
+ offload features of newer <acronym>CPU</acronym>s.
+ <application>bhyve</application> also avoids emulating
+ compatible hardware for the guest, and instead relies on the
+ paravirtualization drivers, called
+ <literal>VirtIO</literal>.</para>
+
+ <para>Due to the design of <application>bhyve</application>, it is
+ only possible to use <application>bhyve</application> on
+ computers with newer processors that support &intel;
+ <acronym>EPT</acronym> (Extended Page Tables) or &amd;
+ <acronym>RVI</acronym> (Rapid Virtualization Indexing, also know
+ as <acronym>NPT</acronym> or Nested Page Tables). Most newer
+ processors, specifically the &intel; &core; i3/i5/i7 and
+ &intel; &xeon; E3/E5/E7 support this feature, for a
+ complete list of &intel; processors that support
+ <acronym>EPT</acronym> see the <link
+ xlink:href="http://ark.intel.com/search/advanced?s=t&ExtendedPageTables=true">&intel;
+ ARK</link>. <acronym>RVI</acronym> is found on the 3rd
+ generation and later of the &amd.opteron; (Barcelona)
+ processors. The easiest way to check for support of
+ <acronym>EPT</acronym> or <acronym>RVI</acronym> on a system is
+ to look for the <literal>POPCNT</literal> processor feature flag
+ on the <literal>Features2</literal> line in
+ <command>dmesg</command> or
+ <filename>/var/run/dmesg.boot</filename>.</para>
+
+ <sect2 xml:id="virtualization-bhyve-prep">
+ <title>Preparing the Host</title>
+
+ <para>The first step to creating a virtual machine in
+ <application>bhyve</application> is configuring the host
+ system. Load the <application>bhyve</application> kernel
+ module called vmm. Create a <filename>tap</filename>
+ interface for the network device in the virtual machine to
+ attach to. Optionally create a bridge interface and add the
+ <filename>tap</filename> interface as well as the physical
+ interface as members to allow the virtual machine to have
+ access to the network.</para>
+
+ <screen>&prompt.root; <userinput>kldload vmm</userinput>
+&prompt.root; <userinput>kldload nmdm</userinput>
+&prompt.root; <userinput>ifconfig <replaceable>tap0</replaceable> create</userinput>
+&prompt.root; <userinput>sysctl net.link.tap.up_on_open=1</userinput>
+net.link.tap.up_on_open: 0 -> 1
+&prompt.root; <userinput>ifconfig <replaceable>bridge0</replaceable> create</userinput>
+&prompt.root; <userinput>ifconfig <replaceable>bridge0</replaceable> addm <replaceable>igb0</replaceable> addm <replaceable>tap0</replaceable></userinput>
+&prompt.root; <userinput>ifconfig <replaceable>bridge0</replaceable> up</userinput></screen>
+
+ </sect2>
+
+ <sect2 xml:id="virtualization-bhyve-freebsd">
+ <title>Creating a FreeBSD Guest</title>
+
+ <para>Create a file to use as the virtual disk for the guest
+ machine.</para>
+
+ <screen>&prompt.root; <userinput>truncate -s <replaceable>16G</replaceable> <filename>guest.img</filename></userinput></screen>
+
+ <para>Download an installation image of &os; to install:</para>
+
+ <screen>&prompt.root; <userinput>fetch <replaceable>ftp://ftp.freebsd.org/pub/FreeBSD/ISO-IMAGES-amd64/10.0/FreeBSD-10.0-RELEASE-amd64-bootonly.iso</replaceable></userinput>;
+FreeBSD-10.0-RELEASE-amd64-bootonly.iso 100% of 209 MB 570 kBps 06m17s</screen>
+
+ <para>&os; comes with an example script for running a virtual
+ machine in <application>bhyve</application>. The script will
+ start the virtual machine and run it in a loop, so it will
+ automatically restart if it crashes. The script takes a
+ number of options to control the configuration of the machine.
+ <option>-c</option> controls the number of virtual CPUs.
+ <option>-m</option> limits the amount of memory available to
+ the guest. <option>-t</option> defines which
+ <filename>tap</filename> device to use. <option>-d</option>
+ indicates which disk image to use. <option>-i</option> tells
+ <application>bhyve</application> to boot from the CD image
+ instead of the disk, and <option>-I</option> defines which CD
+ image to use. Finally the last parameter is the name of the
+ virtual machine, used to track the running machines. Start
+ the virtual machine in installation mode:</para>
+
+ <screen>&prompt.root; <userinput>sh <filename>/usr/share/examples/bhyve/vmrun.sh</filename> -c <replaceable>4</replaceable> -m <replaceable>1024M</replaceable> -t tap0 -d <filename>guest.img</filename> -i -I <filename>FreeBSD-10.0-RELEASE-amd64-bootonly.iso</filename> <replaceable>guestname</replaceable></userinput></screen>
+
+ <para>The system will boot and start the installer. After
+ installing a system in the virtual machine, when the system
+ asks about dropping in to a shell at the end of the
+ installation, choose <guibutton>Yes</guibutton>. A small
+ change needs to be made to make the system start with a serial
+ console. Edit <filename>/etc/ttys</filename> and replace the
+ existing <literal>console</literal> line with:</para>
+
+ <programlisting>console "/usr/libexec/getty std.9600" xterm on secure</programlisting>
+
+ <para>Reboot the virtual machine. Rebooting the virtual machine
+ causes <application>bhyve</application> to exit. However the
+ <filename>vmrun.sh</filename> script runs
+ <command>bhyve</command> in a loop and will automatically
+ restart it. When this happens, choose the reboot option from
+ the boot loader menu, and this will escape the loop. Now the
+ guest can be started from the virtual disk:</para>
+
+ <screen>&prompt.root; <userinput>sh <filename>/usr/share/examples/bhyve/vmrun.sh</filename> -c <replaceable>4</replaceable> -m <replaceable>1024M</replaceable> -t tap0 -d <filename>guest.img</filename> <replaceable>guestname</replaceable></userinput></screen>
+ </sect2>
+
+ <sect2 xml:id="virtualization-bhyve-linux">
+ <title>Creating a &linux; Guest</title>
+
+ <note><para><application>bhyve</application> requires
+ <package>sysutils/grub2-bhyve</package> in order to boot
+ operating systems other than &os;.</para></note>
+
+ <para>Create a file to use as the virtual disk for the guest
+ machine.</para>
+
+ <screen>&prompt.root; <userinput>truncate -s <replaceable>16G</replaceable> <filename>linux.img</filename></userinput></screen>
+
+ <para>Starting a virtual machine with
+ <application>bhyve</application> is a two step process. First
+ a kernel must be loaded, then the guest can be started.
+ <package>sysutils/grub2-bhyve</package> is used to load the
+ &linux; kernel. Create a <filename>device.map</filename> that
+ <application>grub</application> will use to map the virtual
+ devices to the files on the host system:</para>
+
+ <programlisting>(hd0) ./linux.img
+(cd0) ./somelinux.iso</programlisting>
+
+ <para>Use <package>sysutils/grub2-bhyve</package> to load the
+ &linux; kernel from the <acronym>ISO</acronym> image:</para>
+
+ <screen>&prompt.root; <userinput>grub-bhyve -m <filename>device.map</filename> -r cd0 -M <replaceable>1024M</replaceable> <replaceable>linuxguest</replaceable></userinput></screen>
+
+ <para>This will start grub. If the installation CD contains a
+ <filename>grub.cfg</filename> then a menu will be displayed.
+ If not, the <literal>vmlinuz</literal> and
+ <literal>initrd</literal> files must be located and loaded
+ manually:</para>
+
+ <screen>grub> <userinput>ls</userinput>
+(hd0) (cd0) (cd0,msdos1) (host)
+grub> <userinput>ls (cd0)/isolinux</userinput>
+boot.cat boot.msg grub.conf initrd.img isolinux.bin isolinux.cfg memtest
+splash.jpg TRANS.TBL vesamenu.c32 vmlinuz
+grub> <userinput>linux (cd0)/isolinux/vmlinuz</userinput>
+grub> <userinput>initrd (cd0)/isolinux/initrd.img</userinput>
+grub> <userinput>boot</userinput></screen>
+
+ <para>Now that the &linux; kernel is loaded, the guest can be
+ started:</para>
+
+ <screen>&prompt.root; <userinput>bhyve -AI -H -P \
+-s 0:0,hostbridge \
+-s 1:0,lpc \
+-s 2:0,virtio-net,tap1 \
+-s 3:0,virtio-blk,./linux.img \
+-s 4:0,ahci-cd,./somelinux.iso \
+-l com1,stdio \
+-c <replaceable>4</replaceable> -m <replaceable>1024M</replaceable> <replaceable>linuxguest</replaceable></userinput></screen>
+
+ <para>The system will boot and start the installer. After
+ installing a system in the virtual machine, reboot the virtual
+ machine. This will cause <application>bhyve</application> to
+ exit. The instance of the virtual machine needs to be
+ destroyed before it can be started again:</para>
+
+ <screen>&prompt.root; <userinput>bhyvectl --destroy --vm=<replaceable>linuxguest</replaceable></userinput></screen>
+
+ <para>Now the guest can be started directly from the virtual
+ disk. Load the kernel:</para>
+
+ <screen>&prompt.root; <userinput>grub-bhyve -m <filename>device.map</filename> -r hd0,msdos1 -M <replaceable>1024M</replaceable> <replaceable>linuxguest</replaceable></userinput>
+grub> <userinput>ls</userinput>
+(hd0) (hd0,msdos2) (hd0,msdos1) (cd0) (cd0,msdos1) (host)
+(lvm/VolGroup-lv_swap) (lvm/VolGroup-lv_root)
+grub> <userinput>ls (hd0,msdos1)/</userinput>
+lost+found/ grub/ efi/ System.map-2.6.32-431.el6.x86_64 config-2.6.32-431.el6.x
+86_64 symvers-2.6.32-431.el6.x86_64.gz vmlinuz-2.6.32-431.el6.x86_64
+initramfs-2.6.32-431.el6.x86_64.img
+grub> <userinput>linux (hd0,msdos1)/vmlinuz-2.6.32-431.el6.x86_64 root=/dev/mapper/VolGroup-lv_root</userinput>
+grub> <userinput>initrd (hd0,msdos1)/initramfs-2.6.32-431.el6.x86_64.img</userinput>
+grub> <userinput>boot</userinput></screen>
+
+ <para>Boot the virtual machine:</para>
+
+ <screen>&prompt.root; <userinput>bhyve -AI -H -P \
+-s 0:0,hostbridge \
+-s 1:0,lpc \
+-s 2:0,virtio-net,tap1 \
+-s 3:0,virtio-blk,./linux.img \
+-l com1,stdio \
+-c <replaceable>4</replaceable> -m <replaceable>1024M</replaceable> <replaceable>linuxguest</replaceable></userinput></screen>
+
+ <para>&linux; will now boot in the virtual machine and
+ eventually present you with the login prompt. Login and use
+ the virtual machine. When you are finished, reboot the
+ virtual machine to exit <application>bhyve</application>.
+ Destroy the virtual machine instance:</para>
+
+ <screen>&prompt.root; <userinput>bhyvectl --destroy --vm=<replaceable>linuxguest</replaceable></userinput></screen>
+ </sect2>
+
+ <sect2 xml:id="virtualization-bhyve-nmdm">
+ <title>Virtual Machines Consoles</title>
+
+ <para>It is advantageous to wrap the
+ <application>bhyve</application> console in a session
+ management tool such as <package>sysutils/tmux</package> or
+ <package>sysutils/screen</package> in order to detach and
+ reattach to the console. It is also possible to have the
+ console of <application>bhyve</application> be a null modem
+ device that can be accessed with <command>cu</command>. Load
+ the <filename>nmdm</filename> kernel module, and replace
+ <option>-l com1,stdio</option> with
+ <option>-l com1,/dev/nmdm0A</option>. The
+ <filename>/dev/nmdm</filename> devices are created
+ automatically as needed, each is a pair,
+ <filename>/dev/nmdm1A</filename> and
+ <filename>/dev/nmdm1B</filename> corresponding to the two ends
+ of the null modem cable. See &man.nmdm.4; for more
+ information.</para>
+
+ <screen>&prompt.root; <userinput>bhyve -AI -H -P \
+-s 0:0,hostbridge \
+-s 1:0,lpc \
+-s 2:0,virtio-net,tap1 \
+-s 3:0,virtio-blk,./linux.img \
+-l com1,<replaceable>/dev/nmdm0A</replaceable> \
+-c <replaceable>4</replaceable> -m <replaceable>1024M</replaceable> <replaceable>linuxguest</replaceable></userinput>
+&prompt.root; <userinput>cu -l /dev/nmdm0B -s 9600</userinput>
+Connected
+
+Ubuntu 13.10 handbook ttyS0
+
+handbook login:</screen>
+
+ </sect2>
+
+ <sect2 xml:id="virtualization-bhyve-managing">
+ <title>Managing Virtual Machines</title>
+
+ <para>A device node is created in <filename
+ role="directory">/dev/vmm</filename> for each virtual
+ machine. This allows the administrator to easily see a list
+ of the running virtual machines:</para>
+
+ <screen>&prompt.root; <userinput>ls -al /dev/vmm</userinput>
+total 1
+dr-xr-xr-x 2 root wheel 512 Mar 17 12:19 ./
+dr-xr-xr-x 14 root wheel 512 Mar 17 06:38 ../
+crw------- 1 root wheel 0x1a2 Mar 17 12:20 guestname
+crw------- 1 root wheel 0x19f Mar 17 12:19 linuxguest
+crw------- 1 root wheel 0x1a1 Mar 17 12:19 otherguest</screen>
+
+ <para>Virtual machines can be destroyed using
+ <command>bhyvectl</command>:</para>
+
+ <screen>&prompt.root; bhyvectl --destroy --vm=guestname</screen>
+ </sect2>
+
+ <sect2 xml:id="virtualization-bhyve-onboot">
+ <title>Persistent Configuration</title>
+
+ <para>In order to make the system able to start
+ <application>bhyve</application> guests at boot time, the
+ following configurations must be made in the specified
+ files:</para>
+
+ <procedure>
+ <step>
+ <title><filename>/etc/sysctl.conf</filename></title>
+
+ <programlisting>net.link.tap.up_on_open=1</programlisting>
+ </step>
+
+ <step>
+ <title><filename>/boot/loader.conf</filename></title>
+
+ <programlisting>vmm_load="YES"
+nmdm_load="YES"
+if_bridge_load="YES"
+if_tap_load="YES"</programlisting>
+ </step>
+
+ <step>
+ <title><filename>/etc/rc.conf</filename></title>
+
+ <programlisting>cloned_interfaces="bridge0 tap0"
+ifconfig_bridge0="addm igb0 addm tap0"</programlisting>
+ </step>
+ </procedure>
+ </sect2>
<!--
Note: There is no working/end-user ready Xen support for FreeBSD as of 07-2010.
Hide all information regarding Xen under FreeBSD.
>Release-Note:
>Audit-Trail:
>Unformatted:
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201405020127.s421Rk5L039399>