File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.28: download - view: text, annotated - select for diffs
Wed Dec 24 14:35:23 2014 UTC (4 years, 11 months ago) by gdt
Branches: MAIN
CVS tags: HEAD
Add skeleton for generic domU section.

Add section about VPS domU, where you don't control dom0.

Merge link section.

    1: Introduction
    2: ============
    3: 
    4: [![[Xen
    5: screenshot]](http://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](../../gallery/in-Action/hubertf-xen.png)
    6: 
    7: Xen is a virtual machine monitor or hypervisor for x86 hardware
    8: (i686-class or higher), which supports running multiple guest
    9: operating systems on a single physical machine.  With Xen, one uses
   10: the Xen kernel to control the CPU, memory and console, a dom0
   11: operating system which mediates access to other hardware (e.g., disks,
   12: network, USB), and one or more domU operating systems which operate in
   13: an unprivileged virtualized environment.  IO requests from the domU
   14: systems are forwarded by the hypervisor (Xen) to the dom0 to be
   15: fulfilled.
   16: 
   17: Xen supports two styles of guests.  The original is Para-Virtualized
   18: (PV) which means that the guest OS does not attempt to access hardware
   19: directly, but instead makes hypercalls to the hypervisor.  This is
   20: analogous to a user-space program making system calls.  (The dom0
   21: operating system uses PV calls for some functions, such as updating
   22: memory mapping page tables, but has direct hardware access for disk
   23: and network.)   PV guests must be specifically coded for Xen.
   24: 
   25: The more recent style is HVM, which means that the guest does not have
   26: code for Xen and need not be aware that it is running under Xen.
   27: Attempts to access hardware registers are trapped and emulated.  This
   28: style is less efficient but can run unmodified guests.
   29: 
   30: Generally any amd64 machine will work with Xen and PV guests.  For
   31: HVM guests, the VT or VMX cpu feature (Intel) or SVM/HVM/VT (amd64)
   32: is needed; "cpuctl identify 0" will show this.  Xen 4.2 is the last
   33: version for support for using i386 as a host.  TODO: Clean up and
   34: check the above features.
   35: 
   36: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
   37: The dom0 can start one or more domUs.  (Booting is explained in detail
   38: in the dom0 section.)
   39: 
   40: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
   41: and that Xen kernels and tools are available in pkgsrc.  This HOWTO
   42: attempts to address both the case of running a NetBSD dom0 on hardware
   43: and running domUs under it (NetBSD and other), and also running NetBSD
   44: as a domU in a VPS.
   45: 
   46: Some versions of Xen support "PCI passthrough", which means that
   47: specific PCI devices can be made available to a specific domU instead
   48: of the dom0.  This can be useful to let a domU run X11, or access some
   49: network interface or other peripheral.
   50: 
   51: Prerequisites
   52: -------------
   53: 
   54: Installing NetBSD/Xen is not extremely difficult, but it is more
   55: complex than a normal installation of NetBSD.
   56: In general, this HOWTO is occasionally overly restrictive about how
   57: things must be done, guiding the reader to stay on the established
   58: path when there are no known good reasons to stray.
   59: 
   60: This HOWTO presumes a basic familiarity with the Xen system
   61: architecture.  This HOWTO presumes familiarity with installing NetBSD
   62: on i386/amd64 hardware and installing software from pkgsrc.
   63: See also the [Xen website](http://www.xenproject.org/).
   64: 
   65: History
   66: -------
   67: 
   68: NetBSD used to support Xen2; this has been removed.
   69: 
   70: Before NetBSD's native bootloader could support Xen, the use of
   71: grub was recommended.  If necessary, see the
   72: [old grub information](/ports/xen/howto-grub/).
   73: 
   74: Versions of Xen and NetBSD
   75: ==========================
   76: 
   77: Most of the installation concepts and instructions are independent
   78: of Xen version and NetBSD version.  This section gives advice on
   79: which version to choose.  Versions not in pkgsrc and older unsupported
   80: versions of NetBSD are intentionally ignored.
   81: 
   82: Xen
   83: ---
   84: 
   85: In NetBSD, xen is provided in pkgsrc, via matching pairs of packages
   86: xenkernel and xentools.  We will refer only to the kernel versions,
   87: but note that both packages must be installed together and must have
   88: matching versions.
   89: 
   90: xenkernel3 and xenkernel33 provide Xen 3.1 and 3.3.  These no longer
   91: receive security patches and should not be used.  Xen 3.1 supports PCI
   92: passthrough.
   93: 
   94: xenkernel41 provides Xen 4.1.  This is no longer maintained by Xen,
   95: but as of 2014-12 receives backported security patches.  It is a
   96: reasonable although trailing-edge choice.
   97: 
   98: xenkernel42 provides Xen 4.2.  This is maintained by Xen, but old as
   99: of 2014-12.
  100: 
  101: Ideally newer versions of Xen will be added to pkgsrc.
  102: 
  103: Note that NetBSD support is called XEN3.  It works with 3.1 through
  104: 4.2 because the hypercall interface has been stable.
  105: 
  106: Xen command program
  107: -------------------
  108: 
  109: Early Xen used a program called "xm" to manipulate the system from the
  110: dom0.  Starting in 4.1, a replacement program with similar behavior
  111: called "xl" is provided.  In 4.2 and later, "xl" is preferred.  4.4 is
  112: the last version that has "xm".
  113: 
  114: NetBSD
  115: ------
  116: 
  117: The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
  118: reasonable choices, with more or less the same considerations for
  119: non-Xen use.  Therefore, netbsd-6 is recommended as the stable version
  120: of the most recent release.
  121: 
  122: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
  123: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
  124: need SMP; the lack of support is really a problem when using a dom0 as
  125: a normal computer.)
  126: 
  127: Architecture
  128: ------------
  129: 
  130: Xen is basically amd64 only at this point.  One can either run i386
  131: domains or amd64 domains.  If running i386, PAE versions are required,
  132: for both dom0 and domU.  These versions are built by default in NetBSD
  133: releases.  While i386 dom0 works fine, amd64 is recommended as more
  134: normal.  (Note that emacs (at least) fails if run on i386 with PAE when
  135: built without, and vice versa, presumably due to bugs in the undump
  136: code.)
  137: 
  138: Recommendation
  139: --------------
  140: 
  141: Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
  142: xl, the NetBSD 6 stable branch, and to use amd64 as the dom0.  Either
  143: the i386 or amd64 of NetBSD may be used as domUs.
  144: 
  145: NetBSD as a dom0
  146: ================
  147: 
  148: NetBSD can be used as a dom0 and works very well.  The following
  149: sections address installation, updating NetBSD, and updating Xen.
  150: Note that it doesn't make sense to talk about installing a dom0 OS
  151: without also installing Xen itself.  We first address installing
  152: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
  153: NetBSD install to a dom0 install by just changing the kernel and boot
  154: configuration.
  155: 
  156: Styles of dom0 operation
  157: ------------------------
  158: 
  159: There are two basic ways to use Xen.  The traditional method is for
  160: the dom0 to do absolutely nothing other than providing support to some
  161: number of domUs.  Such a system was probably installed for the sole
  162: purpose of hosting domUs, and sits in a server room on a UPS.
  163: 
  164: The other way is to put Xen under a normal-usage computer, so that the
  165: dom0 is what the computer would have been without Xen, perhaps a
  166: desktop or laptop.  Then, one can run domUs at will.  Purists will
  167: deride this as less secure than the previous approach, and for a
  168: computer whose purpose is to run domUs, they are right.  But Xen and a
  169: dom0 (without domUs) is not meaingfully less secure than the same
  170: things running without Xen.  One can boot Xen or boot regular NetBSD
  171: alternately with little problems, simply refraining from starting the
  172: Xen daemons when not running Xen.
  173: 
  174: Note that NetBSD as dom0 does not support multiple CPUs.  This will
  175: limit the performance of the Xen/dom0 workstation approach.
  176: 
  177: Installation of NetBSD
  178: ----------------------
  179: 
  180: First,
  181: [install NetBSD/amd64](/guide/inst/)
  182: just as you would if you were not using Xen.
  183: However, the partitioning approach is very important.
  184: 
  185: If you want to use RAIDframe for the dom0, there are no special issues
  186: for Xen.  Typically one provides RAID storage for the dom0, and the
  187: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
  188: over a RAID1 header to find /boot from a filesystem within a RAID
  189: partition; this is no different when booting Xen.
  190: 
  191: There are 4 styles of providing backing storage for the virtual disks
  192: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN,
  193: 
  194: With raw partitions, one has a disklabel (or gpt) partition sized for
  195: each virtual disk to be used by the domU.  (If you are able to predict
  196: how domU usage will evolve, please add an explanation to the HOWTO.
  197: Seriously, needs tend to change over time.)
  198: 
  199: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
  200: for domU disks.  This is almost as efficient as raw disk partitions
  201: and more flexible.  Hence raw disk partitions should typically not
  202: be used.
  203: 
  204: One can use files in the dom0 filesystem, typically created by dd'ing
  205: /dev/zero to create a specific size.  This is somewhat less efficient,
  206: but very convenient, as one can cp the files for backup, or move them
  207: between dom0 hosts.
  208: 
  209: Finally, in theory one can place the files backing the domU disks in a
  210: SAN.  (This is an invitation for someone who has done this to add a
  211: HOWTO page.)
  212: 
  213: Installation of Xen
  214: -------------------
  215: 
  216: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
  217: pkgsrc (or another matching pair).
  218: See [the pkgsrc
  219: documentation](http://www.NetBSD.org/docs/pkgsrc/) for help with pkgsrc.
  220: 
  221: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
  222: recent versions have HVM support integrated in the main xentools
  223: package.  It is entirely reasonable to run only PV guests.
  224: 
  225: Next you need to install the selected Xen kernel itself, which is
  226: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
  227: For debugging, one may copy xen-debug.gz; this is conceptually similar
  228: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
  229: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
  230: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
  231: of a NetBSD build.  Both xen and NetBSD may be left compressed.  (If
  232: using i386, use releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.)
  233: 
  234: In a dom0 kernel, kernfs is mandatory for xend to comunicate with the
  235: kernel, so ensure that /kern is in fstab.
  236: 
  237: Because you already installed NetBSD, you have a working boot setup
  238: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
  239: beginning of your root filesystem, /boot present, and likely
  240: /boot.cfg.  (If not, fix before continuing!)
  241: 
  242: See boot.cfg(5) for an example.  The basic line is
  243: 
  244: "menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M"
  245: 
  246: which specifies that the dom0 should have 256M, leaving the rest to be
  247: allocated for domUs.
  248: 
  249: As with non-Xen systems, you should have a line to boot /netbsd (a
  250: kernel that works without Xen) and fallback versions of the non-Xen
  251: kernel, Xen, and the dom0 kernel.
  252: 
  253: The [HowTo on Installing into
  254: RAID-1](http://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
  255: explains how to set up booting a dom0 with Xen using grub with
  256: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
  257: boot.)
  258: 
  259: Configuring Xen
  260: ---------------
  261: 
  262: Now, you have a system that will boot Xen and the dom0 kernel, and
  263: just run the dom0 kernel.  There will be no domUs, and none can be
  264: started because you still have to configure the dom0 tools.
  265: 
  266: For 3.3 (and probably 3.1), add to rc.conf (but note that you should
  267: have installed 4.2):
  268:   xend=YES
  269:   xenbackendd=YES
  270: 
  271: For 4.1 and 4.2, add to rc.conf:
  272:   xend=YES
  273:   xencommons=YES
  274: 
  275: Note that xend is for supporting "xm", and should only be used if
  276: you plan on using "xm".  Do NOT enable xend if you plan on using
  277: "xl" as it will cause problems.
  278: 
  279: Updating NetBSD in a dom0
  280: -------------------------
  281: 
  282: This is just like updating NetBSD on bare hardware, assuming the new
  283: version supports the version of Xen you are running.  Generally, one
  284: replaces the kernel and reboots, and then overlays userland binaries
  285: and adjusts /etc.
  286: 
  287: Note that one must update both the non-Xen kernel typically used for
  288: rescue purposes and the DOM0 kernel used with Xen.
  289: 
  290: To convert from grub to /boot, install an mbr bootblock with fdisk,
  291: bootxx_ with installboot, /boot and /boot.cfg.  This really should be
  292: no different than completely reinstalling boot blocks on a non-Xen
  293: system.
  294: 
  295: Updating Xen versions
  296: ---------------------
  297: 
  298: Updating Xen is conceptually not difficult, but can run into all the
  299: issues found when installing Xen.  Assuming migration from 4.1 to 4.2,
  300: remove the xenkernel41 and xentools41 packages and install the
  301: xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz to /.
  302: 
  303: Ensure that the contents of /etc/rc.d/xen* are correct.  Enable the
  304: correct set of daemons.  Ensure that the domU config files are valid
  305: for the new version.
  306: 
  307: 
  308: Unprivileged domains (domU)
  309: ===========================
  310: 
  311: This section describes general concepts about domUs.  It does not
  312: address specific domU operating systems or how to install them.
  313: 
  314: Provided Resources for PV domains
  315: ---------------------------------
  316: 
  317: TODO: Explain that domUs get cpu, memory, disk and network.
  318: Explain that randomness can be an issue.
  319: 
  320: Virtual disks
  321: -------------
  322: 
  323: TODO: Explain how to set up files for vnd and that one should write all zeros to preallocate.
  324: TODO: Explain in what NetBSD versions sparse vnd files do and don't work.
  325: 
  326: Virtual Networking
  327: ------------------
  328: 
  329: TODO: explain xvif concept, and that it's general.
  330: 
  331: There are two normal styles: bridging and NAT.
  332: 
  333: With bridging, the domU perceives itself to be on the same network as
  334: the dom0.  For server virtualization, this is usually best.
  335: 
  336: With NAT, the domU perceives itself to be behind a NAT running on the
  337: dom0.  This is often appropriate when running Xen on a workstation.
  338: 
  339: One can construct arbitrary other configurations, but there is no
  340: script support.
  341: 
  342: Config files
  343: ------------
  344: 
  345: TODO: give example config files.   Use both lvm and vnd.
  346: 
  347: TODO: explain the mess with 3 arguments for disks and how to cope (0x1).
  348: 
  349: Starting domains
  350: ----------------
  351: 
  352: TODO: Explain "xm start" and "xl start".  Explain rc.d/xendomains.
  353: 
  354: TODO: Explain why 4.1 rc.d/xendomains has xl, when one should use xm
  355: on 4.1.
  356: 
  357: Creating specific unprivileged domains (domU)
  358: =============================================
  359: 
  360: Creating domUs is almost entirely independent of operating system.  We
  361: first explain NetBSD, and then differences for Linux and Solaris.
  362: 
  363: Creating an unprivileged NetBSD domain (domU)
  364: ---------------------------------------------
  365: 
  366: Once you have *domain0* running, you need to start the xen tool daemon
  367: (`/usr/pkg/share/examples/rc.d/xend start`) and the xen backend daemon
  368: (`/usr/pkg/share/examples/rc.d/xenbackendd start` for Xen3\*,
  369: `/usr/pkg/share/examples/rc.d/xencommons start` for Xen4.\*). Make sure
  370: that `/dev/xencons` and `/dev/xenevt` exist before starting `xend`. You
  371: can create them with this command:
  372: 
  373:     # cd /dev && sh MAKEDEV xen
  374: 
  375: xend will write logs to `/var/log/xend.log` and
  376: `/var/log/xend-debug.log`. You can then control xen with the xm tool.
  377: 'xm list' will show something like:
  378: 
  379:     # xm list
  380:     Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  381:     Domain-0           0       64    0  r----     58.1
  382: 
  383: 'xm create' allows you to create a new domain. It uses a config file in
  384: PKG\_SYSCONFDIR for its parameters. By default, this file will be in
  385: `/usr/pkg/etc/xen/`. On creation, a kernel has to be specified, which
  386: will be executed in the new domain (this kernel is in the *domain0* file
  387: system, not on the new domain virtual disk; but please note, you should
  388: install the same kernel into *domainU* as `/netbsd` in order to make
  389: your system tools, like savecore(8), work). A suitable kernel is
  390: provided as part of the i386 and amd64 binary sets: XEN3\_DOMU.
  391: 
  392: Here is an /usr/pkg/etc/xen/nbsd example config file:
  393: 
  394:     #  -*- mode: python; -*-
  395:     #============================================================================
  396:     # Python defaults setup for 'xm create'.
  397:     # Edit this file to reflect the configuration of your system.
  398:     #============================================================================
  399: 
  400:     #----------------------------------------------------------------------------
  401:     # Kernel image file. This kernel will be loaded in the new domain.
  402:     kernel = "/home/bouyer/netbsd-XEN3_DOMU"
  403:     #kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  404: 
  405:     # Memory allocation (in megabytes) for the new domain.
  406:     memory = 128
  407: 
  408:     # A handy name for your new domain. This will appear in 'xm list',
  409:     # and you can use this as parameters for xm in place of the domain
  410:     # number. All domains must have different names.
  411:     #
  412:     name = "nbsd"
  413: 
  414:     # The number of virtual CPUs this domain has.
  415:     #
  416:     vcpus = 1
  417: 
  418:     #----------------------------------------------------------------------------
  419:     # Define network interfaces for the new domain.
  420: 
  421:     # Number of network interfaces (must be at least 1). Default is 1.
  422:     nics = 1
  423: 
  424:     # Define MAC and/or bridge for the network interfaces.
  425:     #
  426:     # The MAC address specified in ``mac'' is the one used for the interface
  427:     # in the new domain. The interface in domain0 will use this address XOR'd
  428:     # with 00:00:00:01:00:00 (i.e. aa:00:00:51:02:f0 in our example). Random
  429:     # MACs are assigned if not given.
  430:     #
  431:     # ``bridge'' is a required parameter, which will be passed to the
  432:     # vif-script called by xend(8) when a new domain is created to configure
  433:     # the new xvif interface in domain0.
  434:     #
  435:     # In this example, the xvif is added to bridge0, which should have been
  436:     # set up prior to the new domain being created -- either in the
  437:     # ``network'' script or using a /etc/ifconfig.bridge0 file.
  438:     #
  439:     vif = [ 'mac=aa:00:00:50:02:f0, bridge=bridge0' ]
  440: 
  441:     #----------------------------------------------------------------------------
  442:     # Define the disk devices you want the domain to have access to, and
  443:     # what you want them accessible as.
  444:     #
  445:     # Each disk entry is of the form:
  446:     #
  447:     #   phy:DEV,VDEV,MODE
  448:     #
  449:     # where DEV is the device, VDEV is the device name the domain will see,
  450:     # and MODE is r for read-only, w for read-write.  You can also create
  451:     # file-backed domains using disk entries of the form:
  452:     #
  453:     #   file:PATH,VDEV,MODE
  454:     #
  455:     # where PATH is the path to the file used as the virtual disk, and VDEV
  456:     # and MODE have the same meaning as for ``phy'' devices.
  457:     #
  458:     # VDEV doesn't really matter for a NetBSD guest OS (it's just used as an index),
  459:     # but it does for Linux.
  460:     # Worse, the device has to exist in /dev/ of domain0, because xm will
  461:     # try to stat() it. This means that in order to load a Linux guest OS
  462:     # from a NetBSD domain0, you'll have to create /dev/hda1, /dev/hda2, ...
  463:     # on domain0, with the major/minor from Linux :(
  464:     # Alternatively it's possible to specify the device number in hex,
  465:     # e.g. 0x301 for /dev/hda1, 0x302 for /dev/hda2, etc ...
  466: 
  467:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  468:     #disk = [ 'file:/var/xen/nbsd-disk,0x01,w' ]
  469:     #disk = [ 'file:/var/xen/nbsd-disk,0x301,w' ]
  470: 
  471:     #----------------------------------------------------------------------------
  472:     # Set the kernel command line for the new domain.
  473: 
  474:     # Set root device. This one does matter for NetBSD
  475:     root = "xbd0"
  476:     # extra parameters passed to the kernel
  477:     # this is where you can set boot flags like -s, -a, etc ...
  478:     #extra = ""
  479: 
  480:     #----------------------------------------------------------------------------
  481:     # Set according to whether you want the domain restarted when it exits.
  482:     # The default is False.
  483:     #autorestart = True
  484: 
  485:     # end of nbsd config file ====================================================
  486: 
  487: When a new domain is created, xen calls the
  488: `/usr/pkg/etc/xen/vif-bridge` script for each virtual network interface
  489: created in *domain0*. This can be used to automatically configure the
  490: xvif?.? interfaces in *domain0*. In our example, these will be bridged
  491: with the bridge0 device in *domain0*, but the bridge has to exist first.
  492: To do this, create the file `/etc/ifconfig.bridge0` and make it look
  493: like this:
  494: 
  495:     create
  496:     !brconfig $int add ex0 up
  497: 
  498: (replace `ex0` with the name of your physical interface). Then bridge0
  499: will be created on boot. See the bridge(4) man page for details.
  500: 
  501: So, here is a suitable `/usr/pkg/etc/xen/vif-bridge` for xvif?.? (a
  502: working vif-bridge is also provided with xentools20) configuring:
  503: 
  504:     #!/bin/sh
  505:     #============================================================================
  506:     # $NetBSD: howto.mdwn,v 1.28 2014/12/24 14:35:23 gdt Exp $
  507:     #
  508:     # /usr/pkg/etc/xen/vif-bridge
  509:     #
  510:     # Script for configuring a vif in bridged mode with a dom0 interface.
  511:     # The xend(8) daemon calls a vif script when bringing a vif up or down.
  512:     # The script name to use is defined in /usr/pkg/etc/xen/xend-config.sxp
  513:     # in the ``vif-script'' field.
  514:     #
  515:     # Usage: vif-bridge up|down [var=value ...]
  516:     #
  517:     # Actions:
  518:     #    up     Adds the vif interface to the bridge.
  519:     #    down   Removes the vif interface from the bridge.
  520:     #
  521:     # Variables:
  522:     #    domain name of the domain the interface is on (required).
  523:     #    vifq   vif interface name (required).
  524:     #    mac    vif MAC address (required).
  525:     #    bridge bridge to add the vif to (required).
  526:     #
  527:     # Example invocation:
  528:     #
  529:     # vif-bridge up domain=VM1 vif=xvif1.0 mac="ee:14:01:d0:ec:af" bridge=bridge0
  530:     #
  531:     #============================================================================
  532: 
  533:     # Exit if anything goes wrong
  534:     set -e
  535: 
  536:     echo "vif-bridge $*"
  537: 
  538:     # Operation name.
  539:     OP=$1; shift
  540: 
  541:     # Pull variables in args into environment
  542:     for arg ; do export "${arg}" ; done
  543: 
  544:     # Required parameters. Fail if not set.
  545:     domain=${domain:?}
  546:     vif=${vif:?}
  547:     mac=${mac:?}
  548:     bridge=${bridge:?}
  549: 
  550:     # Optional parameters. Set defaults.
  551:     ip=${ip:-''}   # default to null (do nothing)
  552: 
  553:     # Are we going up or down?
  554:     case $OP in
  555:     up) brcmd='add' ;;
  556:     down)   brcmd='delete' ;;
  557:     *)
  558:         echo 'Invalid command: ' $OP
  559:         echo 'Valid commands are: up, down'
  560:         exit 1
  561:         ;;
  562:     esac
  563: 
  564:     # Don't do anything if the bridge is "null".
  565:     if [ "${bridge}" = "null" ] ; then
  566:         exit
  567:     fi
  568: 
  569:     # Don't do anything if the bridge doesn't exist.
  570:     if ! ifconfig -l | grep "${bridge}" >/dev/null; then
  571:         exit
  572:     fi
  573: 
  574:     # Add/remove vif to/from bridge.
  575:     ifconfig x${vif} $OP
  576:     brconfig ${bridge} ${brcmd} x${vif}
  577: 
  578: Now, running
  579: 
  580:     xm create -c /usr/pkg/etc/xen/nbsd
  581: 
  582: should create a domain and load a NetBSD kernel in it. (Note: `-c`
  583: causes xm to connect to the domain's console once created.) The kernel
  584: will try to find its root file system on xbd0 (i.e., wd0e) which hasn't
  585: been created yet. wd0e will be seen as a disk device in the new domain,
  586: so it will be 'sub-partitioned'. We could attach a ccd to wd0e in
  587: *domain0* and partition it, newfs and extract the NetBSD/i386 or amd64
  588: tarballs there, but there's an easier way: load the
  589: `netbsd-INSTALL_XEN3_DOMU` kernel provided in the NetBSD binary sets.
  590: Like other install kernels, it contains a ramdisk with sysinst, so you
  591: can install NetBSD using sysinst on your new domain.
  592: 
  593: If you want to install NetBSD/Xen with a CDROM image, the following line
  594: should be used in the `/usr/pkg/etc/xen/nbsd` file:
  595: 
  596:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  597: 
  598: After booting the domain, the option to install via CDROM may be
  599: selected. The CDROM device should be changed to `xbd1d`.
  600: 
  601: Once done installing, `halt -p` the new domain (don't reboot or halt, it
  602: would reload the INSTALL\_XEN3\_DOMU kernel even if you changed the
  603: config file), switch the config file back to the XEN3\_DOMU kernel, and
  604: start the new domain again. Now it should be able to use `root on xbd0a`
  605: and you should have a second, functional NetBSD system on your xen
  606: installation.
  607: 
  608: When the new domain is booting you'll see some warnings about *wscons*
  609: and the pseudo-terminals. These can be fixed by editing the files
  610: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  611: `/etc/ttys`, except *console*, like this:
  612: 
  613:     console "/usr/libexec/getty Pc"         vt100   on secure
  614:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  615:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  616:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  617:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  618: 
  619: Finally, all screens must be commented out from `/etc/wscons.conf`.
  620: 
  621: It is also desirable to add
  622: 
  623:     powerd=YES
  624: 
  625: in rc.conf. This way, the domain will be properly shut down if
  626: `xm shutdown -R` or `xm shutdown -H` is used on the domain0.
  627: 
  628: Your domain should be now ready to work, enjoy.
  629: 
  630: Creating an unprivileged Linux domain (domU)
  631: --------------------------------------------
  632: 
  633: Creating unprivileged Linux domains isn't much different from
  634: unprivileged NetBSD domains, but there are some details to know.
  635: 
  636: First, the second parameter passed to the disk declaration (the '0x1' in
  637: the example below)
  638: 
  639:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  640: 
  641: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  642: for hda). Linux builds device numbers as: (major \<\< 8 + minor). So,
  643: hda1 which has major 3 and minor 1 on a Linux system will have device
  644: number 0x301. Alternatively, devices names can be used (hda, hdb, ...)
  645: as xentools has a table to map these names to devices numbers. To export
  646: a partition to a Linux guest we can use:
  647: 
  648:     disk = [ 'phy:/dev/wd0e,0x300,w' ]
  649:     root = "/dev/hda1 ro"
  650: 
  651: and it will appear as /dev/hda on the Linux system, and be used as root
  652: partition.
  653: 
  654: To install the Linux system on the partition to be exported to the guest
  655: domain, the following method can be used: install sysutils/e2fsprogs
  656: from pkgsrc. Use mke2fs to format the partition that will be the root
  657: partition of your Linux domain, and mount it. Then copy the files from a
  658: working Linux system, make adjustments in `/etc` (fstab, network
  659: config). It should also be possible to extract binary packages such as
  660: .rpm or .deb directly to the mounted partition using the appropriate
  661: tool, possibly running under NetBSD's Linux emulation. Once the
  662: filesystem has been populated, umount it. If desirable, the filesystem
  663: can be converted to ext3 using tune2fs -j. It should now be possible to
  664: boot the Linux guest domain, using one of the vmlinuz-\*-xenU kernels
  665: available in the Xen binary distribution.
  666: 
  667: To get the linux console right, you need to add:
  668: 
  669:     extra = "xencons=tty1"
  670: 
  671: to your configuration since not all linux distributions auto-attach a
  672: tty to the xen console.
  673: 
  674: Creating an unprivileged Solaris domain (domU)
  675: ----------------------------------------------
  676: 
  677: Download an Opensolaris [release](http://opensolaris.org/os/downloads/)
  678: or [development snapshot](http://genunix.org/) DVD image. Attach the DVD
  679: image to a MAN.VND.4 device. Copy the kernel and ramdisk filesystem
  680: image to your dom0 filesystem.
  681: 
  682:     dom0# mkdir /root/solaris
  683:     dom0# vnconfig vnd0 osol-1002-124-x86.iso
  684:     dom0# mount /dev/vnd0a /mnt
  685: 
  686:     ## for a 64-bit guest
  687:     dom0# cp /mnt/boot/amd64/x86.microroot /root/solaris
  688:     dom0# cp /mnt/platform/i86xpv/kernel/amd64/unix /root/solaris
  689: 
  690:     ## for a 32-bit guest
  691:     dom0# cp /mnt/boot/x86.microroot /root/solaris
  692:     dom0# cp /mnt/platform/i86xpv/kernel/unix /root/solaris
  693: 
  694:     dom0# umount /mnt
  695:           
  696: 
  697: Keep the MAN.VND.4 configured. For some reason the boot process stalls
  698: unless the DVD image is attached to the guest as a "phy" device. Create
  699: an initial configuration file with the following contents. Substitute
  700: */dev/wd0k* with an empty partition at least 8 GB large.
  701: 
  702:     memory = 640
  703:     name = 'solaris'
  704:     disk = [ 'phy:/dev/wd0k,0,w' ]
  705:     disk += [ 'phy:/dev/vnd0d,6:cdrom,r' ]
  706:     vif = [ 'bridge=bridge0' ]
  707:     kernel = '/root/solaris/unix'
  708:     ramdisk = '/root/solaris/x86.microroot'
  709:     # for a 64-bit guest
  710:     extra = '/platform/i86xpv/kernel/amd64/unix - nowin -B install_media=cdrom'
  711:     # for a 32-bit guest
  712:     #extra = '/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'
  713:           
  714: 
  715: Start the guest.
  716: 
  717:     dom0# xm create -c solaris.cfg
  718:     Started domain solaris
  719:                           v3.3.2 chgset 'unavailable'
  720:     SunOS Release 5.11 Version snv_124 64-bit
  721:     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  722:     Use is subject to license terms.
  723:     Hostname: opensolaris
  724:     Remounting root read/write
  725:     Probing for device nodes ...
  726:     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  727:     Preparing live image for use
  728:     Done mounting Live image
  729:           
  730: 
  731: Make sure the network is configured. Note that it can take a minute for
  732: the xnf0 interface to appear.
  733: 
  734:     opensolaris console login: jack
  735:     Password: jack
  736:     Sun Microsystems Inc.   SunOS 5.11      snv_124 November 2008
  737:     jack@opensolaris:~$ pfexec sh
  738:     sh-3.2# ifconfig -a
  739:     sh-3.2# exit
  740:           
  741: 
  742: Set a password for VNC and start the VNC server which provides the X11
  743: display where the installation program runs.
  744: 
  745:     jack@opensolaris:~$ vncpasswd
  746:     Password: solaris
  747:     Verify: solaris
  748:     jack@opensolaris:~$ cp .Xclients .vnc/xstartup
  749:     jack@opensolaris:~$ vncserver :1
  750:           
  751: 
  752: From a remote machine connect to the VNC server. Use `ifconfig xnf0` on
  753: the guest to find the correct IP address to use.
  754: 
  755:     remote$ vncviewer 172.18.2.99:1
  756:           
  757: 
  758: It is also possible to launch the installation on a remote X11 display.
  759: 
  760:     jack@opensolaris:~$ export DISPLAY=172.18.1.1:0
  761:     jack@opensolaris:~$ pfexec gui-install
  762:            
  763: 
  764: After the GUI installation is complete you will be asked to reboot.
  765: Before that you need to determine the ZFS ID for the new boot filesystem
  766: and update the configuration file accordingly. Return to the guest
  767: console.
  768: 
  769:     jack@opensolaris:~$ pfexec zdb -vvv rpool | grep bootfs
  770:                     bootfs = 43
  771:     ^C
  772:     jack@opensolaris:~$
  773:            
  774: 
  775: The final configuration file should look like this. Note in particular
  776: the last line.
  777: 
  778:     memory = 640
  779:     name = 'solaris'
  780:     disk = [ 'phy:/dev/wd0k,0,w' ]
  781:     vif = [ 'bridge=bridge0' ]
  782:     kernel = '/root/solaris/unix'
  783:     ramdisk = '/root/solaris/x86.microroot'
  784:     extra = '/platform/i86xpv/kernel/amd64/unix -B zfs-bootfs=rpool/43,bootpath="/xpvd/xdf@0:a"'
  785:            
  786: 
  787: Restart the guest to verify it works correctly.
  788: 
  789:     dom0# xm destroy solaris
  790:     dom0# xm create -c solaris.cfg
  791:     Using config file "./solaris.cfg".
  792:     v3.3.2 chgset 'unavailable'
  793:     Started domain solaris
  794:     SunOS Release 5.11 Version snv_124 64-bit
  795:     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  796:     Use is subject to license terms.
  797:     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  798:     Hostname: osol
  799:     Configuring devices.
  800:     Loading smf(5) service descriptions: 160/160
  801:     svccfg import warnings. See /var/svc/log/system-manifest-import:default.log .
  802:     Reading ZFS config: done.
  803:     Mounting ZFS filesystems: (6/6)
  804:     Creating new rsa public/private host key pair
  805:     Creating new dsa public/private host key pair
  806: 
  807:     osol console login:
  808:            
  809: 
  810: Using PCI devices in guest domains
  811: ----------------------------------
  812: 
  813: The domain0 can give other domains access to selected PCI devices. This
  814: can allow, for example, a non-privileged domain to have access to a
  815: physical network interface or disk controller. However, keep in mind
  816: that giving a domain access to a PCI device most likely will give the
  817: domain read/write access to the whole physical memory, as PCs don't have
  818: an IOMMU to restrict memory access to DMA-capable device. Also, it's not
  819: possible to export ISA devices to non-domain0 domains (which means that
  820: the primary VGA adapter can't be exported. A guest domain trying to
  821: access the VGA registers will panic).
  822: 
  823: This functionality is only available in NetBSD-5.1 (and later) domain0
  824: and domU. If the domain0 is NetBSD, it has to be running Xen 3.1, as
  825: support has not been ported to later versions at this time.
  826: 
  827: For a PCI device to be exported to a domU, is has to be attached to the
  828: `pciback` driver in domain0. Devices passed to the domain0 via the
  829: pciback.hide boot parameter will attach to `pciback` instead of the
  830: usual driver. The list of devices is specified as `(bus:dev.func)`,
  831: where bus and dev are 2-digit hexadecimal numbers, and func a
  832: single-digit number:
  833: 
  834:     pciback.hide=(00:0a.0)(00:06.0)
  835: 
  836: pciback devices should show up in the domain0's boot messages, and the
  837: devices should be listed in the `/kern/xen/pci` directory.
  838: 
  839: PCI devices to be exported to a domU are listed in the `pci` array of
  840: the domU's config file, with the format `'0000:bus:dev.func'`
  841: 
  842:     pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  843: 
  844: In the domU an `xpci` device will show up, to which one or more pci
  845: busses will attach. Then the PCI drivers will attach to PCI busses as
  846: usual. Note that the default NetBSD DOMU kernels do not have `xpci` or
  847: any PCI drivers built in by default; you have to build your own kernel
  848: to use PCI devices in a domU. Here's a kernel config example:
  849: 
  850:     include         "arch/i386/conf/XEN3_DOMU"
  851:     #include         "arch/i386/conf/XENU"           # in NetBSD 3.0
  852: 
  853:     # Add support for PCI busses to the XEN3_DOMU kernel
  854:     xpci* at xenbus ?
  855:     pci* at xpci ?
  856: 
  857:     # Now add PCI and related devices to be used by this domain
  858:     # USB Controller and Devices
  859: 
  860:     # PCI USB controllers
  861:     uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  862: 
  863:     # USB bus support
  864:     usb*    at uhci?
  865: 
  866:     # USB Hubs
  867:     uhub*   at usb?
  868:     uhub*   at uhub? port ? configuration ? interface ?
  869: 
  870:     # USB Mass Storage
  871:     umass*  at uhub? port ? configuration ? interface ?
  872:     wd*     at umass?
  873:     # SCSI controllers
  874:     ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  875: 
  876:     # SCSI bus support (for both ahc and umass)
  877:     scsibus* at scsi?
  878: 
  879:     # SCSI devices
  880:     sd*     at scsibus? target ? lun ?      # SCSI disk drives
  881:     cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  882: 
  883: 
  884: NetBSD as a domU in a VPS
  885: =========================
  886: 
  887: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
  888: hardware.  This section explains how to deal with Xen in a domU as a
  889: virtual private server where you do not control or have access to the
  890: dom0.
  891: 
  892: TODO: Perhaps reference panix, prmgr, amazon as interesting examples.
  893: 
  894: TODO: Somewhere, discuss pvgrub and py-grub to load the domU kernel
  895: from the domU filesystem.

CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb