File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.49: download - view: text, annotated - select for diffs
Fri Dec 26 20:25:19 2014 UTC (4 years, 10 months ago) by gdt
Branches: MAIN
CVS tags: HEAD
redo netbsd domU instructions.

clean up Linux domU from pandoc conversion.

    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.  In
   31: theory i386 computers without amd64 support can be used for Xen <=
   32: 4.2, but we have no recent reports of this working (this is a hint).
   33: For HVM guests, the VT or VMX cpu feature (Intel) or SVM/HVM/VT
   34: (amd64) is needed; "cpuctl identify 0" will show this.  TODO: Clean up
   35: and check the above features.
   36: 
   37: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
   38: The dom0 can start one or more domUs.  (Booting is explained in detail
   39: in the dom0 section.)
   40: 
   41: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
   42: and that Xen kernels and tools are available in pkgsrc.  This HOWTO
   43: attempts to address both the case of running a NetBSD dom0 on hardware
   44: and running domUs under it (NetBSD and other), and also running NetBSD
   45: as a domU in a VPS.
   46: 
   47: Some versions of Xen support "PCI passthrough", which means that
   48: specific PCI devices can be made available to a specific domU instead
   49: of the dom0.  This can be useful to let a domU run X11, or access some
   50: network interface or other peripheral.
   51: 
   52: Prerequisites
   53: -------------
   54: 
   55: Installing NetBSD/Xen is not extremely difficult, but it is more
   56: complex than a normal installation of NetBSD.
   57: In general, this HOWTO is occasionally overly restrictive about how
   58: things must be done, guiding the reader to stay on the established
   59: path when there are no known good reasons to stray.
   60: 
   61: This HOWTO presumes a basic familiarity with the Xen system
   62: architecture.  This HOWTO presumes familiarity with installing NetBSD
   63: on i386/amd64 hardware and installing software from pkgsrc.
   64: See also the [Xen website](http://www.xenproject.org/).
   65: 
   66: History
   67: -------
   68: 
   69: NetBSD used to support Xen2; this has been removed.
   70: 
   71: Before NetBSD's native bootloader could support Xen, the use of
   72: grub was recommended.  If necessary, see the
   73: [old grub information](/ports/xen/howto-grub/).
   74: 
   75: Versions of Xen and NetBSD
   76: ==========================
   77: 
   78: Most of the installation concepts and instructions are independent
   79: of Xen version and NetBSD version.  This section gives advice on
   80: which version to choose.  Versions not in pkgsrc and older unsupported
   81: versions of NetBSD are intentionally ignored.
   82: 
   83: Xen
   84: ---
   85: 
   86: In NetBSD, xen is provided in pkgsrc, via matching pairs of packages
   87: xenkernel and xentools.  We will refer only to the kernel versions,
   88: but note that both packages must be installed together and must have
   89: matching versions.
   90: 
   91: xenkernel3 and xenkernel33 provide Xen 3.1 and 3.3.  These no longer
   92: receive security patches and should not be used.  Xen 3.1 supports PCI
   93: passthrough.  Xen 3.1 supports non-PAE on i386.
   94: 
   95: xenkernel41 provides Xen 4.1.  This is no longer maintained by Xen,
   96: but as of 2014-12 receives backported security patches.  It is a
   97: reasonable although trailing-edge choice.
   98: 
   99: xenkernel42 provides Xen 4.2.  This is maintained by Xen, but old as
  100: of 2014-12.
  101: 
  102: Ideally newer versions of Xen will be added to pkgsrc.
  103: 
  104: Note that NetBSD support is called XEN3.  It works with 3.1 through
  105: 4.2 because the hypercall interface has been stable.
  106: 
  107: Xen command program
  108: -------------------
  109: 
  110: Early Xen used a program called "xm" to manipulate the system from the
  111: dom0.  Starting in 4.1, a replacement program with similar behavior
  112: called "xl" is provided.  In 4.2 and later, "xl" is preferred.  4.4 is
  113: the last version that has "xm".
  114: 
  115: NetBSD
  116: ------
  117: 
  118: The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
  119: reasonable choices, with more or less the same considerations for
  120: non-Xen use.  Therefore, netbsd-6 is recommended as the stable version
  121: of the most recent release for production use.  For those wanting to
  122: learn Xen or without production stability concerns, netbsd-7 is likely
  123: most appropriate.
  124: 
  125: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
  126: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
  127: need SMP; the lack of support is really a problem when using a dom0 as
  128: a normal computer.)
  129: 
  130: Architecture
  131: ------------
  132: 
  133: Xen itself can run on i386 or amd64 machines.  (Practically, almost
  134: any computer where one would want to run Xen supports amd64.)  If
  135: using an i386 NetBSD kernel for the dom0, PAE is required (PAE
  136: versions are built by default).  While i386 dom0 works fine, amd64 is
  137: recommended as more normal.
  138: 
  139: Xen 4.2 is the last version to support i386 as a host.  TODO: Clarify
  140: if this is about the CPU having to be amd64, or about the dom0 kernel
  141: having to be amd64.
  142: 
  143: One can then run i386 domUs and amd64 domUs, in any combination.  If
  144: running an i386 NetBSD kernel as a domU, the PAE version is required.
  145: (Note that emacs (at least) fails if run on i386 with PAE when built
  146: without, and vice versa, presumably due to bugs in the undump code.)
  147: 
  148: Recommendation
  149: --------------
  150: 
  151: Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
  152: xl, the NetBSD 6 stable branch, and to use an amd64 kernel as the
  153: dom0.  Either the i386 or amd64 of NetBSD may be used as domUs.
  154: 
  155: Build problems
  156: --------------
  157: 
  158: Ideally, all versions of Xen in pkgsrc would build on all versions of
  159: NetBSD on both i386 and amd64.  However, that isn't the case.  Besides
  160: aging code and aging compilers, qemu (included in xentools for HVM
  161: support) is difficult to build.  The following are known to fail:
  162: 
  163:         xenkernel3 netbsd-6 i386
  164:         xentools42 netbsd-6 i386 
  165: 
  166: The following are known to work:
  167: 
  168:         xenkernel41 netbsd-5 amd64
  169:         xentools41 netbsd-5 amd64
  170:         xenkernel41 netbsd-6 i386
  171:         xentools41 netbsd-6 i386
  172: 
  173: NetBSD as a dom0
  174: ================
  175: 
  176: NetBSD can be used as a dom0 and works very well.  The following
  177: sections address installation, updating NetBSD, and updating Xen.
  178: Note that it doesn't make sense to talk about installing a dom0 OS
  179: without also installing Xen itself.  We first address installing
  180: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
  181: NetBSD install to a dom0 install by just changing the kernel and boot
  182: configuration.
  183: 
  184: For experimenting with Xen, a machine with as little as 1G of RAM and
  185: 100G of disk can work.  For running many domUs in productions, far
  186: more will be needed.
  187: 
  188: Styles of dom0 operation
  189: ------------------------
  190: 
  191: There are two basic ways to use Xen.  The traditional method is for
  192: the dom0 to do absolutely nothing other than providing support to some
  193: number of domUs.  Such a system was probably installed for the sole
  194: purpose of hosting domUs, and sits in a server room on a UPS.
  195: 
  196: The other way is to put Xen under a normal-usage computer, so that the
  197: dom0 is what the computer would have been without Xen, perhaps a
  198: desktop or laptop.  Then, one can run domUs at will.  Purists will
  199: deride this as less secure than the previous approach, and for a
  200: computer whose purpose is to run domUs, they are right.  But Xen and a
  201: dom0 (without domUs) is not meaingfully less secure than the same
  202: things running without Xen.  One can boot Xen or boot regular NetBSD
  203: alternately with little problems, simply refraining from starting the
  204: Xen daemons when not running Xen.
  205: 
  206: Note that NetBSD as dom0 does not support multiple CPUs.  This will
  207: limit the performance of the Xen/dom0 workstation approach.
  208: 
  209: Installation of NetBSD
  210: ----------------------
  211: 
  212: First,
  213: [install NetBSD/amd64](/guide/inst/)
  214: just as you would if you were not using Xen.
  215: However, the partitioning approach is very important.
  216: 
  217: If you want to use RAIDframe for the dom0, there are no special issues
  218: for Xen.  Typically one provides RAID storage for the dom0, and the
  219: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
  220: over a RAID1 header to find /boot from a filesystem within a RAID
  221: partition; this is no different when booting Xen.
  222: 
  223: There are 4 styles of providing backing storage for the virtual disks
  224: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN,
  225: 
  226: With raw partitions, one has a disklabel (or gpt) partition sized for
  227: each virtual disk to be used by the domU.  (If you are able to predict
  228: how domU usage will evolve, please add an explanation to the HOWTO.
  229: Seriously, needs tend to change over time.)
  230: 
  231: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
  232: for domU disks.  This is almost as efficient as raw disk partitions
  233: and more flexible.  Hence raw disk partitions should typically not
  234: be used.
  235: 
  236: One can use files in the dom0 filesystem, typically created by dd'ing
  237: /dev/zero to create a specific size.  This is somewhat less efficient,
  238: but very convenient, as one can cp the files for backup, or move them
  239: between dom0 hosts.
  240: 
  241: Finally, in theory one can place the files backing the domU disks in a
  242: SAN.  (This is an invitation for someone who has done this to add a
  243: HOWTO page.)
  244: 
  245: Installation of Xen
  246: -------------------
  247: 
  248: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
  249: pkgsrc (or another matching pair).
  250: See [the pkgsrc
  251: documentation](http://www.NetBSD.org/docs/pkgsrc/) for help with pkgsrc.
  252: 
  253: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
  254: recent versions have HVM support integrated in the main xentools
  255: package.  It is entirely reasonable to run only PV guests.
  256: 
  257: Next you need to install the selected Xen kernel itself, which is
  258: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
  259: For debugging, one may copy xen-debug.gz; this is conceptually similar
  260: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
  261: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
  262: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
  263: of a NetBSD build.  Both xen and NetBSD may be left compressed.  (If
  264: using i386, use releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.)
  265: 
  266: With Xen as the kernel, you must provide a dom0 NetBSD kernel to be
  267: used as a module; place this in /.  Suitable kernels are provided in
  268: releasedir/binary/kernel:
  269: 
  270:         i386 XEN3_DOM0
  271:         i386 XEN3PAE_DOM0
  272: 	amd64 XEN3_DOM0
  273: 
  274: The first one is only for use with Xen 3.1 and i386-mode Xen (and you
  275: should not do this).  Current Xen always uses PAE on i386, but you
  276: should generally use amd64 for the dom0.  In a dom0 kernel, kernfs is
  277: mandatory for xend to comunicate with the kernel, so ensure that /kern
  278: is in fstab.  TODO: Say this is default, or file a PR and give a
  279: reference.
  280: 
  281: Because you already installed NetBSD, you have a working boot setup
  282: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
  283: beginning of your root filesystem, /boot present, and likely
  284: /boot.cfg.  (If not, fix before continuing!)
  285: 
  286: See boot.cfg(5) for an example.  The basic line is
  287: 
  288:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
  289: 
  290: which specifies that the dom0 should have 256M, leaving the rest to be
  291: allocated for domUs.  In an attempt to add performance, one can also
  292: add
  293: 
  294:         dom0_max_vcpus=1 dom0_vcpus_pin
  295: 
  296: to force only one vcpu to be provided (since NetBSD dom0 can't use
  297: more) and to pin that vcpu to a physical cpu.  TODO: benchmark this.
  298: 
  299: As with non-Xen systems, you should have a line to boot /netbsd (a
  300: kernel that works without Xen) and fallback versions of the non-Xen
  301: kernel, Xen, and the dom0 kernel.
  302: 
  303: The [HowTo on Installing into
  304: RAID-1](http://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
  305: explains how to set up booting a dom0 with Xen using grub with
  306: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
  307: boot.)
  308: 
  309: Configuring Xen
  310: ---------------
  311: 
  312: Now, you have a system that will boot Xen and the dom0 kernel, and
  313: just run the dom0 kernel.  There will be no domUs, and none can be
  314: started because you still have to configure the dom0 tools.  The
  315: daemons which should be run vary with Xen version and with whether one
  316: is using xm or xl.  Note that xend is for supporting "xm", and should
  317: only be used if you plan on using "xm".  Do NOT enable xend if you
  318: plan on using "xl" as it will cause problems.
  319: 
  320: The installation of NetBSD should already have created devices for xen
  321: (xencons, xenevt), but if they are not present, create them:
  322: 
  323:         cd /dev && sh MAKEDEV xen
  324: 
  325: TODO: Give 3.1 advice (or remove it from pkgsrc).
  326: 
  327: For 3.3 (and thus xm), add to rc.conf (but note that you should have
  328: installed 4.1 or 4.2):
  329: 
  330:         xend=YES
  331:         xenbackendd=YES
  332: 
  333: For 4.1 (and thus xm; xl is believed not to work well), add to rc.conf:
  334: 
  335:         xend=YES
  336:         xencommons=YES
  337: 
  338: TODO: Explain why if xm is preferred on 4.1, rc.d/xendomains has xl.
  339: Or fix the package.
  340: 
  341: For 4.2 with xm, add to rc.conf
  342: 
  343:         xend=YES
  344:         xencommons=YES
  345: 
  346: For 4.2 with xl (preferred), add to rc.conf:
  347: 
  348:         TODO: explain if there is a xend replacement
  349:         xencommons=YES
  350: 
  351: TODO: Recommend for/against xen-watchdog.
  352: 
  353: After you have configured the daemons and either started them or
  354: rebooted, run the following (or use xl) to inspect Xen's boot
  355: messages, available resources, and running domains:
  356: 
  357:         # xm dmesg
  358: 	[xen's boot info]
  359:         # xm info
  360: 	[available memory, etc.]
  361:         # xm list
  362:         Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  363:         Domain-0           0       64    0  r----     58.1
  364: 
  365: anita (for testing NetBSD)
  366: --------------------------
  367: 
  368: With the setup so far, one should be able to run anita (see
  369: pkgsrc/sysutils/py-anita) to test NetBSD releases, by doing (as root,
  370: because anita must create a domU):
  371: 
  372:         anita --vmm=xm test file:///usr/obj/i386/
  373: 
  374: Alternatively, one can use --vmm=xl to use xl-based domU creation instead.
  375: TODO: check this.
  376: 
  377: Xen-specific NetBSD issues
  378: --------------------------
  379: 
  380: There are (at least) two additional things different about NetBSD as a
  381: dom0 kernel compared to hardware.
  382: 
  383: One is that modules are not usable in DOM0 kernels, so one must
  384: compile in what's needed.  It's not really that modules cannot work,
  385: but that modules must be built for XEN3_DOM0 because some of the
  386: defines change and the normal module builds don't do this.  Basically,
  387: enabling Xen changes the kernel ABI, and the module build system
  388: doesn't cope with this.
  389: 
  390: The other difference is that XEN3_DOM0 does not have exactly the same
  391: options as GENERIC.  While it is debatable whether or not this is a
  392: bug, users should be aware of this and can simply add missing config
  393: items if desired.
  394: 
  395: Updating NetBSD in a dom0
  396: -------------------------
  397: 
  398: This is just like updating NetBSD on bare hardware, assuming the new
  399: version supports the version of Xen you are running.  Generally, one
  400: replaces the kernel and reboots, and then overlays userland binaries
  401: and adjusts /etc.
  402: 
  403: Note that one must update both the non-Xen kernel typically used for
  404: rescue purposes and the DOM0 kernel used with Xen.
  405: 
  406: To convert from grub to /boot, install an mbr bootblock with fdisk,
  407: bootxx_ with installboot, /boot and /boot.cfg.  This really should be
  408: no different than completely reinstalling boot blocks on a non-Xen
  409: system.
  410: 
  411: Updating Xen versions
  412: ---------------------
  413: 
  414: Updating Xen is conceptually not difficult, but can run into all the
  415: issues found when installing Xen.  Assuming migration from 4.1 to 4.2,
  416: remove the xenkernel41 and xentools41 packages and install the
  417: xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz to /.
  418: 
  419: Ensure that the contents of /etc/rc.d/xen* are correct.  Enable the
  420: correct set of daemons.  Ensure that the domU config files are valid
  421: for the new version.
  422: 
  423: 
  424: Unprivileged domains (domU)
  425: ===========================
  426: 
  427: This section describes general concepts about domUs.  It does not
  428: address specific domU operating systems or how to install them.  The
  429: config files for domUs are typically in /usr/pkg/etc/xen, and are
  430: typically named so that the file anme, domU name and the domU's host
  431: name match.
  432: 
  433: The domU is provided with cpu and memory by Xen, configured by the
  434: dom0.  The domU is provided with disk and network by the dom0,
  435: mediated by Xen, and configured in the dom0.
  436: 
  437: Entropy in domUs can be an issue; physical disks and network are on
  438: the dom0.  NetBSD's /dev/random system works, but is often challenged.
  439: 
  440: Config files
  441: ------------
  442: 
  443: There is no good order to present config files and the concepts
  444: surrounding what is being configured.  We first show an example config
  445: file, and then in the various sections give details.
  446: 
  447: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
  448: for a large number of well-commented examples, mostly for running
  449: GNU/Linux.
  450: 
  451: The following is an example minimal domain configuration file
  452: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
  453: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
  454: i386 domU).  The domU serves as a network file server.
  455: 
  456:         # -*- mode: python; -*-
  457: 
  458:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
  459:         memory = 1024
  460:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
  461:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
  462:                  'file:/n0/xen/foo-wd1,0x1,w' ]
  463: 
  464: The domain will have the same name as the file.  The kernel has the
  465: host/domU name in it, so that on the dom0 one can update the various
  466: domUs independently.  The vif line causes an interface to be provided,
  467: with a specific mac address (do not reuse MAC addresses!), in bridge
  468: mode.  Two disks are provided, and they are both writable; the bits
  469: are stored in files and Xen attaches them to a vnd(4) device in the
  470: dom0 on domain creation.  The system treates xbd0 as the boot device
  471: without needing explicit configuration.
  472: 
  473: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
  474: that "xm create" takes the name of a config file, while other commands
  475: take the name of a domain.  To create the domain, connect to the
  476: console, create the domain while attaching the console, shutdown the
  477: domain, and see if it has finished stopping, do (or xl with Xen >=
  478: 4.2):
  479: 
  480:         xm create foo
  481:         xm console foo
  482:         xm create -c foo
  483:         xm shutdown foo
  484: 	xm list
  485: 
  486: Typing ^] will exit the console session.  Shutting down a domain is
  487: equivalent to pushing the power button; a NetBSD domU will receive a
  488: power-press event and do a clean shutdown.  Shutting down the dom0
  489: will trigger controlled shutdowns of all configured domUs.
  490: 
  491: domU kernels
  492: ------------
  493: 
  494: On a physical computer, the BIOS reads sector 0, and a chain of boot
  495: loaders finds and loads a kernel.  Normally this comes from the root
  496: filesystem.  With Xen domUs, the process is totally different.  The
  497: normal path is for the domU kernel to be a file in the dom0's
  498: filesystem.  At the request of the dom0, Xen loads that kernel into a
  499: new domU instance and starts execution.  While domU kernels can be
  500: anyplace, reasonable places to store domU kernels on the dom0 are in /
  501: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
  502: config files), or in /u0/xen (where the vdisks are).
  503: 
  504: See the VPS section near the end for discussion of alternate ways to
  505: obtain domU kernels.
  506: 
  507: CPU and memory
  508: --------------
  509: 
  510: A domain is provided with some number of vcpus, less than the number
  511: of cpus seen by the hypervisor.  (For a dom0, this is controlled by
  512: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
  513: from the config file by the "vcpus = N" directive.
  514: 
  515: A domain is provided with memory; this is controlled in the config
  516: file by "memory = N" (in megabytes).  In the straightforward case, the
  517: sum of the the memory allocated to the dom0 and all domUs must be less
  518: than the available memory.
  519: 
  520: Xen also provides a "balloon" driver, which can be used to let domains
  521: use more memory temporarily.  TODO: Explain better, and explain how
  522: well it works with NetBSD.
  523: 
  524: Virtual disks
  525: -------------
  526: 
  527: With the file/vnd style, typically one creates a directory,
  528: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
  529: domUs.  Then, for each domU disk, one writes zeros to a file that then
  530: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
  531: for the first virtual disk for the domU called foo.  Writing zeros to
  532: the file serves two purposes.  One is that preallocating the contents
  533: improves performance.  The other is that vnd on sparse files has
  534: failed to work.  TODO: give working/notworking NetBSD versions for
  535: sparse vnd.  Note that the use of file/vnd for Xen is not really
  536: different than creating a file-backed virtual disk for some other
  537: purpose, except that xentools handles the vnconfig commands.  To
  538: create an empty 4G virtual disk, simply do
  539: 
  540:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
  541: 
  542: With the lvm style, one creates logical devices.  They are then used
  543: similarly to vnds.  TODO: Add an example with lvm.
  544: 
  545: In domU config files, the disks are defined as a sequence of 3-tuples.
  546: The first element is "method:/path/to/disk".  Common methods are
  547: "file:" for file-backed vnd. and "phy:" for something that is already
  548: a (TODO: character or block) device.
  549: 
  550: The second element is an artifact of how virtual disks are passed to
  551: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
  552: are given a device name to associate with the disk, and values like
  553: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
  554: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
  555: second argument.  The name given is converted to a major/minor by
  556: calling stat(2) on the name in /dev and this is passed to the domU.
  557: In the general case, the dom0 and domU can be different operating
  558: systems, and it is an unwarranted assumption that they have consistent
  559: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
  560: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
  561: for the second works fine and avoids this issue.  For a GNU/Linux
  562: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
  563: /dev/hda1.
  564: 
  565: The third element is "w" for writable disks, and "r" for read-only
  566: disks.
  567: 
  568: Virtual Networking
  569: ------------------
  570: 
  571: Xen provides virtual ethernets, each of which connects the dom0 and a
  572: domU.  For each virtual network, there is an interface "xvifN.M" in
  573: the dom0, and in domU index N, a matching interface xennetM (NetBSD
  574: name).  The interfaces behave as if there is an Ethernet with two
  575: adaptors connected.  From this primitive, one can construct various
  576: configurations.  We focus on two common and useful cases for which
  577: there are existing scripts: bridging and NAT.
  578: 
  579: With bridging (in the example above), the domU perceives itself to be
  580: on the same network as the dom0.  For server virtualization, this is
  581: usually best.  Bridging is accomplished by creating a bridge(4) device
  582: and adding the dom0's physical interface and the various xvifN.0
  583: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
  584: config file.  The bridge must be set up already in the dom0; an
  585: example /etc/ifconfig.bridge0 is:
  586: 
  587:         create
  588:         up
  589:         !brconfig bridge0 add wm0
  590: 
  591: With NAT, the domU perceives itself to be behind a NAT running on the
  592: dom0.  This is often appropriate when running Xen on a workstation.
  593: TODO: NAT appears to be configured by "vif = [ '' ]".
  594: 
  595: The MAC address specified is the one used for the interface in the new
  596: domain.  The interface in domain0 will use this address XOR'd with
  597: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
  598: 
  599: Sizing domains
  600: --------------
  601: 
  602: Modern x86 hardware has vast amounts of resources.  However, many
  603: virtual servers can function just fine on far less.  A system with
  604: 256M of RAM and a 4G disk can be a reasonable choice.  Note that it is
  605: far easier to adjust virtual resources than physical ones.  For
  606: memory, it's just a config file edit and a reboot.  For disk, one can
  607: create a new file and vnconfig it (or lvm), and then dump/restore,
  608: just like updating physical disks, but without having to be there and
  609: without those pesky connectors.
  610: 
  611: Starting domains automatically
  612: ------------------------------
  613: 
  614: To start domains foo at bar at boot and shut them down cleanly on dom0
  615: shutdown, in rc.conf add:
  616: 
  617:         xendomains="foo bar"
  618: 
  619: TODO: Explain why 4.1 rc.d/xendomains has xl, when one should use xm
  620: on 4.1.  Or fix the xentools41 package to have xm
  621: 
  622: Creating specific unprivileged domains (domU)
  623: =============================================
  624: 
  625: Creating domUs is almost entirely independent of operating system.  We
  626: have already presented the basics of config files.  Note that you must
  627: have already completed the dom0 setup so that "xl list" (or "xm list")
  628: works.
  629: 
  630: Creating an unprivileged NetBSD domain (domU)
  631: ---------------------------------------------
  632: 
  633: See the earlier config file, and adjust memory.  Decide on how much
  634: storage you will provide, and prepare it (file or lvm).
  635: 
  636: While the kernel will be obtained from the dom0 filesystem, the same
  637: file should be present in the domU as /netbsd so that tools like
  638: savecore(8) can work.   (This is helpful but not necessary.)
  639: 
  640: The kernel must be specifically for Xen and for use as a domU.  The
  641: i386 and amd64 provide the following kernels:
  642: 
  643:         i386 XEN3_DOMU
  644:         i386 XEN3PAE_DOMU
  645: 	amd64 XEN3_DOMU
  646: 
  647: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
  648: use the PAE version of the i386 kernel.
  649: 
  650: This will boot NetBSD, but this is not that useful if the disk is
  651: empty.  One approach is to unpack sets onto the disk outside of xen
  652: (by mounting it, just as you would prepare a physical disk for a
  653: system you can't run the installer on).
  654: 
  655: A second approach is to run an INSTALL kernel, which has a miniroot
  656: and can load sets from the network.  To do this, copy the INSTALL
  657: kernel to / and change the kernel line in the config file to:
  658: 
  659:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  660: 
  661: Then, start the domain as "xl create -c configname".
  662: 
  663: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
  664: line should be used in the config file.
  665: 
  666:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  667: 
  668: After booting the domain, the option to install via CDROM may be
  669: selected.  The CDROM device should be changed to `xbd1d`.
  670: 
  671: Once done installing, "halt -p" the new domain (don't reboot or halt,
  672: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
  673: config file), switch the config file back to the XEN3_DOMU kernel,
  674: and start the new domain again. Now it should be able to use "root on
  675: xbd0a" and you should have a, functional NetBSD domU.
  676: 
  677: TODO: check if this is still accurate.
  678: When the new domain is booting you'll see some warnings about *wscons*
  679: and the pseudo-terminals. These can be fixed by editing the files
  680: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  681: `/etc/ttys`, except *console*, like this:
  682: 
  683:     console "/usr/libexec/getty Pc"         vt100   on secure
  684:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  685:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  686:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  687:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  688: 
  689: Finally, all screens must be commented out from `/etc/wscons.conf`.
  690: 
  691: It is also desirable to add
  692: 
  693:         powerd=YES
  694: 
  695: in rc.conf. This way, the domain will be properly shut down if
  696: `xm shutdown -R` or `xm shutdown -H` is used on the domain0.
  697: 
  698: Your domain should be now ready to work, enjoy.
  699: 
  700: Creating an unprivileged Linux domain (domU)
  701: --------------------------------------------
  702: 
  703: Creating unprivileged Linux domains isn't much different from
  704: unprivileged NetBSD domains, but there are some details to know.
  705: 
  706: First, the second parameter passed to the disk declaration (the '0x1' in
  707: the example below)
  708: 
  709:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  710: 
  711: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  712: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
  713: So, hda1 which has major 3 and minor 1 on a Linux system will have
  714: device number 0x301.  Alternatively, devices names can be used (hda,
  715: hdb, ...)  as xentools has a table to map these names to devices
  716: numbers.  To export a partition to a Linux guest we can use:
  717: 
  718:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
  719:         root = "/dev/hda1 ro"
  720: 
  721: and it will appear as /dev/hda on the Linux system, and be used as root
  722: partition.
  723: 
  724: To install the Linux system on the partition to be exported to the
  725: guest domain, the following method can be used: install
  726: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
  727: that will be the root partition of your Linux domain, and mount it.
  728: Then copy the files from a working Linux system, make adjustments in
  729: `/etc` (fstab, network config).  It should also be possible to extract
  730: binary packages such as .rpm or .deb directly to the mounted partition
  731: using the appropriate tool, possibly running under NetBSD's Linux
  732: emulation.  Once the filesystem has been populated, umount it.  If
  733: desirable, the filesystem can be converted to ext3 using tune2fs -j.
  734: It should now be possible to boot the Linux guest domain, using one of
  735: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
  736: 
  737: To get the linux console right, you need to add:
  738: 
  739:     extra = "xencons=tty1"
  740: 
  741: to your configuration since not all linux distributions auto-attach a
  742: tty to the xen console.
  743: 
  744: Creating an unprivileged Solaris domain (domU)
  745: ----------------------------------------------
  746: 
  747: Download an Opensolaris [release](http://opensolaris.org/os/downloads/)
  748: or [development snapshot](http://genunix.org/) DVD image. Attach the DVD
  749: image to a MAN.VND.4 device. Copy the kernel and ramdisk filesystem
  750: image to your dom0 filesystem.
  751: 
  752:     dom0# mkdir /root/solaris
  753:     dom0# vnconfig vnd0 osol-1002-124-x86.iso
  754:     dom0# mount /dev/vnd0a /mnt
  755: 
  756:     ## for a 64-bit guest
  757:     dom0# cp /mnt/boot/amd64/x86.microroot /root/solaris
  758:     dom0# cp /mnt/platform/i86xpv/kernel/amd64/unix /root/solaris
  759: 
  760:     ## for a 32-bit guest
  761:     dom0# cp /mnt/boot/x86.microroot /root/solaris
  762:     dom0# cp /mnt/platform/i86xpv/kernel/unix /root/solaris
  763: 
  764:     dom0# umount /mnt
  765:           
  766: 
  767: Keep the MAN.VND.4 configured. For some reason the boot process stalls
  768: unless the DVD image is attached to the guest as a "phy" device. Create
  769: an initial configuration file with the following contents. Substitute
  770: */dev/wd0k* with an empty partition at least 8 GB large.
  771: 
  772:     memory = 640
  773:     name = 'solaris'
  774:     disk = [ 'phy:/dev/wd0k,0,w' ]
  775:     disk += [ 'phy:/dev/vnd0d,6:cdrom,r' ]
  776:     vif = [ 'bridge=bridge0' ]
  777:     kernel = '/root/solaris/unix'
  778:     ramdisk = '/root/solaris/x86.microroot'
  779:     # for a 64-bit guest
  780:     extra = '/platform/i86xpv/kernel/amd64/unix - nowin -B install_media=cdrom'
  781:     # for a 32-bit guest
  782:     #extra = '/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'
  783:           
  784: 
  785: Start the guest.
  786: 
  787:     dom0# xm create -c solaris.cfg
  788:     Started domain solaris
  789:                           v3.3.2 chgset 'unavailable'
  790:     SunOS Release 5.11 Version snv_124 64-bit
  791:     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  792:     Use is subject to license terms.
  793:     Hostname: opensolaris
  794:     Remounting root read/write
  795:     Probing for device nodes ...
  796:     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  797:     Preparing live image for use
  798:     Done mounting Live image
  799:           
  800: 
  801: Make sure the network is configured. Note that it can take a minute for
  802: the xnf0 interface to appear.
  803: 
  804:     opensolaris console login: jack
  805:     Password: jack
  806:     Sun Microsystems Inc.   SunOS 5.11      snv_124 November 2008
  807:     jack@opensolaris:~$ pfexec sh
  808:     sh-3.2# ifconfig -a
  809:     sh-3.2# exit
  810:           
  811: 
  812: Set a password for VNC and start the VNC server which provides the X11
  813: display where the installation program runs.
  814: 
  815:     jack@opensolaris:~$ vncpasswd
  816:     Password: solaris
  817:     Verify: solaris
  818:     jack@opensolaris:~$ cp .Xclients .vnc/xstartup
  819:     jack@opensolaris:~$ vncserver :1
  820:           
  821: 
  822: From a remote machine connect to the VNC server. Use `ifconfig xnf0` on
  823: the guest to find the correct IP address to use.
  824: 
  825:     remote$ vncviewer 172.18.2.99:1
  826:           
  827: 
  828: It is also possible to launch the installation on a remote X11 display.
  829: 
  830:     jack@opensolaris:~$ export DISPLAY=172.18.1.1:0
  831:     jack@opensolaris:~$ pfexec gui-install
  832:            
  833: 
  834: After the GUI installation is complete you will be asked to reboot.
  835: Before that you need to determine the ZFS ID for the new boot filesystem
  836: and update the configuration file accordingly. Return to the guest
  837: console.
  838: 
  839:     jack@opensolaris:~$ pfexec zdb -vvv rpool | grep bootfs
  840:                     bootfs = 43
  841:     ^C
  842:     jack@opensolaris:~$
  843:            
  844: 
  845: The final configuration file should look like this. Note in particular
  846: the last line.
  847: 
  848:     memory = 640
  849:     name = 'solaris'
  850:     disk = [ 'phy:/dev/wd0k,0,w' ]
  851:     vif = [ 'bridge=bridge0' ]
  852:     kernel = '/root/solaris/unix'
  853:     ramdisk = '/root/solaris/x86.microroot'
  854:     extra = '/platform/i86xpv/kernel/amd64/unix -B zfs-bootfs=rpool/43,bootpath="/xpvd/xdf@0:a"'
  855:            
  856: 
  857: Restart the guest to verify it works correctly.
  858: 
  859:     dom0# xm destroy solaris
  860:     dom0# xm create -c solaris.cfg
  861:     Using config file "./solaris.cfg".
  862:     v3.3.2 chgset 'unavailable'
  863:     Started domain solaris
  864:     SunOS Release 5.11 Version snv_124 64-bit
  865:     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  866:     Use is subject to license terms.
  867:     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  868:     Hostname: osol
  869:     Configuring devices.
  870:     Loading smf(5) service descriptions: 160/160
  871:     svccfg import warnings. See /var/svc/log/system-manifest-import:default.log .
  872:     Reading ZFS config: done.
  873:     Mounting ZFS filesystems: (6/6)
  874:     Creating new rsa public/private host key pair
  875:     Creating new dsa public/private host key pair
  876: 
  877:     osol console login:
  878:            
  879: 
  880: Using PCI devices in guest domains
  881: ----------------------------------
  882: 
  883: The domain0 can give other domains access to selected PCI devices. This
  884: can allow, for example, a non-privileged domain to have access to a
  885: physical network interface or disk controller. However, keep in mind
  886: that giving a domain access to a PCI device most likely will give the
  887: domain read/write access to the whole physical memory, as PCs don't have
  888: an IOMMU to restrict memory access to DMA-capable device. Also, it's not
  889: possible to export ISA devices to non-domain0 domains (which means that
  890: the primary VGA adapter can't be exported. A guest domain trying to
  891: access the VGA registers will panic).
  892: 
  893: This functionality is only available in NetBSD-5.1 (and later) domain0
  894: and domU. If the domain0 is NetBSD, it has to be running Xen 3.1, as
  895: support has not been ported to later versions at this time.
  896: 
  897: For a PCI device to be exported to a domU, is has to be attached to the
  898: `pciback` driver in domain0. Devices passed to the domain0 via the
  899: pciback.hide boot parameter will attach to `pciback` instead of the
  900: usual driver. The list of devices is specified as `(bus:dev.func)`,
  901: where bus and dev are 2-digit hexadecimal numbers, and func a
  902: single-digit number:
  903: 
  904:     pciback.hide=(00:0a.0)(00:06.0)
  905: 
  906: pciback devices should show up in the domain0's boot messages, and the
  907: devices should be listed in the `/kern/xen/pci` directory.
  908: 
  909: PCI devices to be exported to a domU are listed in the `pci` array of
  910: the domU's config file, with the format `'0000:bus:dev.func'`
  911: 
  912:     pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  913: 
  914: In the domU an `xpci` device will show up, to which one or more pci
  915: busses will attach. Then the PCI drivers will attach to PCI busses as
  916: usual. Note that the default NetBSD DOMU kernels do not have `xpci` or
  917: any PCI drivers built in by default; you have to build your own kernel
  918: to use PCI devices in a domU. Here's a kernel config example:
  919: 
  920:     include         "arch/i386/conf/XEN3_DOMU"
  921:     #include         "arch/i386/conf/XENU"           # in NetBSD 3.0
  922: 
  923:     # Add support for PCI busses to the XEN3_DOMU kernel
  924:     xpci* at xenbus ?
  925:     pci* at xpci ?
  926: 
  927:     # Now add PCI and related devices to be used by this domain
  928:     # USB Controller and Devices
  929: 
  930:     # PCI USB controllers
  931:     uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  932: 
  933:     # USB bus support
  934:     usb*    at uhci?
  935: 
  936:     # USB Hubs
  937:     uhub*   at usb?
  938:     uhub*   at uhub? port ? configuration ? interface ?
  939: 
  940:     # USB Mass Storage
  941:     umass*  at uhub? port ? configuration ? interface ?
  942:     wd*     at umass?
  943:     # SCSI controllers
  944:     ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  945: 
  946:     # SCSI bus support (for both ahc and umass)
  947:     scsibus* at scsi?
  948: 
  949:     # SCSI devices
  950:     sd*     at scsibus? target ? lun ?      # SCSI disk drives
  951:     cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  952: 
  953: 
  954: NetBSD as a domU in a VPS
  955: =========================
  956: 
  957: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
  958: hardware.  This section explains how to deal with Xen in a domU as a
  959: virtual private server where you do not control or have access to the
  960: dom0.
  961: 
  962: TODO: Perhaps reference panix, prmgr, amazon as interesting examples.
  963: 
  964: TODO: Somewhere, discuss pvgrub and py-grub to load the domU kernel
  965: from the domU filesystem.
  966: 
  967: Using npf
  968: ---------
  969: 
  970: In standard kernels, npf is a module, and thus cannot be loadeed in a
  971: DOMU kernel.
  972: 
  973: TODO: explain how to compile npf into a custom kernel, answering:
  974: http://mail-index.netbsd.org/netbsd-users/2014/12/26/msg015576.html

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