File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.154: download - view: text, annotated - select for diffs
Thu Jul 26 16:29:45 2018 UTC (16 months, 2 weeks ago) by maxv
Branches: MAIN
CVS tags: HEAD
Remove references to non-PAE i386.

    1: [[!meta title="Xen HowTo"]]
    2: 
    3: Introduction
    4: ============
    5: 
    6: [![[Xen
    7: screenshot]](https://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](https://www.netbsd.org/gallery/in-Action/hubertf-xen.png)
    8: 
    9: Xen is a hypervisor for x86 hardware, which supports running multiple guest
   10: operating systems on a single physical machine.  Xen is a Type 1 or
   11: bare-metal hypervisor; one uses the Xen kernel to control the CPU,
   12: memory and console, a dom0 operating system which mediates access to
   13: other hardware (e.g., disks, network, USB), and one or more domU
   14: operating systems which operate in an unprivileged virtualized
   15: environment.  IO requests from the domU systems are forwarded by the
   16: Xen hypervisor to the dom0 to be fulfilled.
   17: 
   18: Xen supports different styles of guest:
   19: 
   20: [[!table data="""
   21: Style of guest	|Supported by NetBSD
   22: PV		|Yes (dom0, domU)
   23: HVM		|Yes (domU)
   24: PVHVM		|No
   25: PVH		|No
   26: """]]
   27: 
   28: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
   29: hardware directly, but instead makes hypercalls to the hypervisor; PV
   30: guests must be specifically coded for Xen. In HVM mode, no guest
   31: modification is required; however, hardware support is required, such
   32: as VT-x on Intel CPUs and SVM on AMD CPUs.
   33: 
   34: There are further features for IOMMU virtualization, Intel's VT-d and
   35: AMD's AMD-Vi.  TODO: Explain whether Xen on NetBSD makes use of these
   36: features.  TODO: Review by someone who really understands this.
   37: 
   38: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
   39: The dom0 can start one or more domUs.  (Booting is explained in detail
   40: in the dom0 section.)
   41: 
   42: This HOWTO presumes a basic familiarity with the Xen system
   43: architecture, with installing NetBSD on i386/amd64 hardware, and with
   44: installing software from pkgsrc.  See also the [Xen
   45: website](http://www.xenproject.org/).
   46: 
   47: This HOWTO attempts to address both the case of running a NetBSD dom0
   48: on hardware and running domUs under it (NetBSD and other), and also
   49: running NetBSD as a domU in a VPS.
   50: 
   51: Versions of Xen and NetBSD
   52: ==========================
   53: 
   54: Most of the installation concepts and instructions are independent
   55: of Xen version and NetBSD version.  This section gives advice on
   56: which version to choose.  Versions not in pkgsrc and older unsupported
   57: versions of NetBSD are intentionally ignored.
   58: 
   59: The term "amd64" is used to refer to both the NetBSD port and to the
   60: hardware architecture on which it runs.  Such hardware is generally
   61: made by both Intel and AMD, and common on PC computers.
   62: 
   63: Xen versions
   64: ------------
   65: 
   66: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
   67: xenkernel and xentools.  We will refer only to the kernel versions,
   68: but note that both packages must be installed together and must have
   69: matching versions.
   70: 
   71: Versions available in pkgsrc:
   72: 
   73: [[!table data="""
   74: Xen Version	|Package Name	|Xen CPU Support	|EOL'ed By Upstream
   75: 4.2		|xenkernel42	|32bit, 64bit		|Yes
   76: 4.5		|xenkernel45	|64bit			|Yes
   77: 4.6		|xenkernel46	|64bit			|Partially
   78: 4.8		|xenkernel48	|64bit			|No
   79: 4.11		|xenkernel411	|64bit			|No
   80: """]]
   81: 
   82: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
   83: 
   84: Note: Xen 4.2 was the last version to support 32bit CPUs.
   85: 
   86: Xen command program
   87: -------------------
   88: 
   89: Early Xen used a program called xm to manipulate the system from the
   90: dom0.  Starting in 4.1, a replacement program with similar behavior
   91: called xl is provided, but it does not work well in 4.1.  In 4.2, both
   92: xm and xl work fine.  4.4 is the last version that has xm.
   93: 
   94: You must make a global choice to use xm or xl, because it affects not
   95: only which command you use, but the command used by rc.d scripts
   96: (specifically xendomains) and which daemons should be run.  The
   97: xentools packages provide xl for 4.2 and up.
   98: 
   99: In 4.2, you can choose to use xm by simply changing the ctl_command
  100: variable and setting xend=YES in rc.conf.
  101: 
  102: With xl, virtual devices are configured in parallel, which can cause
  103: problems if they are written assuming serial operation (e.g., updating
  104: firewall rules without explicit locking).  There is now locking for
  105: the provided scripts, which works for normal casses (e.g, file-backed
  106: xbd, where a vnd must be allocated).  But, as of 201612, it has not
  107: been adequately tested for a complex custom setup with a large number
  108: of interfaces.
  109: 
  110: NetBSD versions
  111: ---------------
  112: 
  113: The netbsd-7, netbsd-8, and -current branches are all reasonable
  114: choices, with more or less the same considerations for non-Xen use.
  115: NetBSD 8 is recommended as the stable version of the most recent
  116: release for production use.
  117: 
  118: For developing Xen, netbsd-current may be appropriate.
  119: 
  120: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
  121: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
  122: need SMP for dom0 functions; the lack of support is really a problem
  123: when using a dom0 as a normal computer.)
  124: 
  125: Note: NetBSD support is called XEN3. However, it does support Xen 4,
  126: because the hypercall interface has remained identical.
  127: 
  128: Architecture
  129: ------------
  130: 
  131: Xen itself can run on i386 (Xen < 4.2) or amd64 hardware (all Xen
  132: versions).  Practically, almost any computer where one would want to
  133: run Xen today supports amd64.
  134: 
  135: Xen, the dom0 system, and each domU system can be either i386 or
  136: amd64.  When building a xenkernel package, one obtains an i386 Xen
  137: kernel on an i386 host, and an amd64 Xen kernel on an amd64 host.  If
  138: the Xen kernel is i386, then the dom0 kernel and all domU kernels must
  139: be i386.  With an amd64 Xen kernel, an amd64 dom0 kernel is known to
  140: work, and an i386 dom0 kernel should in theory work.  An amd64
  141: Xen/dom0 is known to support both i386 and amd64 domUs.
  142: 
  143: i386 dom0 and domU kernels must be PAE. PAE kernels are included in
  144: the NetBSD default build.
  145: 
  146: Because of the above, the standard approach is to use an amd64 Xen
  147: kernel and NetBSD/amd64 for the dom0.  For domUs, NetBSD/i386 (PAE) and
  148: NetBSD/amd64 are in widespread use, and there is
  149: little to no Xen-specific reason to prefer one over the other.
  150: 
  151: Note that to use an i386 dom0 with Xen 4.5 or higher, one must build
  152: (or obtain from pre-built packages) an amd64 Xen kernel and install
  153: that on the system.  Almost no one in the NetBSD/Xen community does
  154: this, and the standard, well-tested, approach is to use an amd64 dom0.
  155: 
  156: A [posting on
  157: xen-devel](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html)
  158: explained that PV system call overhead was higher on amd64, and thus
  159: there is some notion that i386 guests are faster.  It goes on to
  160: caution that the total situation is complex and not entirely
  161: understood. On top of that caution, the post is about Linux, not
  162: NetBSD.  TODO: Include link to benchmarks, if someone posts them.
  163: 
  164: NetBSD as a dom0
  165: ================
  166: 
  167: NetBSD can be used as a dom0 and works very well.  The following
  168: sections address installation, updating NetBSD, and updating Xen.
  169: Note that it doesn't make sense to talk about installing a dom0 OS
  170: without also installing Xen itself.  We first address installing
  171: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
  172: NetBSD install to a dom0 install by just changing the kernel and boot
  173: configuration.
  174: 
  175: For experimenting with Xen, a machine with as little as 1G of RAM and
  176: 100G of disk can work.  For running many domUs in productions, far
  177: more will be needed; e.g. 4-8G and 1T of disk is reasonable for a
  178: half-dozen domUs of 512M and 32G each.  Basically, the RAM and disk
  179: have to be bigger than the sum of the RAM/disk needs of the dom0 and
  180: all the domUs.
  181: 
  182: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
  183: 512M it worked reliably.  This does not make sense, but if you see
  184: "not ELF" after Xen boots, try increasing dom0 RAM.
  185: 
  186: Styles of dom0 operation
  187: ------------------------
  188: 
  189: There are two basic ways to use Xen.  The traditional method is for
  190: the dom0 to do absolutely nothing other than providing support to some
  191: number of domUs.  Such a system was probably installed for the sole
  192: purpose of hosting domUs, and sits in a server room on a UPS.
  193: 
  194: The other way is to put Xen under a normal-usage computer, so that the
  195: dom0 is what the computer would have been without Xen, perhaps a
  196: desktop or laptop.  Then, one can run domUs at will.  Purists will
  197: deride this as less secure than the previous approach, and for a
  198: computer whose purpose is to run domUs, they are right.  But Xen and a
  199: dom0 (without domUs) is not meaningfully less secure than the same
  200: things running without Xen.  One can boot Xen or boot regular NetBSD
  201: alternately with little problems, simply refraining from starting the
  202: Xen daemons when not running Xen.
  203: 
  204: Note that NetBSD as dom0 does not support multiple CPUs.  This will
  205: limit the performance of the Xen/dom0 workstation approach.  In theory
  206: the only issue is that the "backend drivers" are not yet MPSAFE:
  207:   https://mail-index.netbsd.org/netbsd-users/2014/08/29/msg015195.html
  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 file system 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 file system, 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).  See [the pkgsrc
  250: documentation](https://www.NetBSD.org/docs/pkgsrc/) for help with
  251: pkgsrc.  Ensure that your packages are recent; the HOWTO does not
  252: contemplate old builds.
  253: 
  254: Next you need to install the selected Xen kernel itself, which is
  255: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
  256: For debugging, one may copy xen-debug.gz; this is conceptually similar
  257: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
  258: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
  259: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
  260: of a NetBSD build.  If using i386, use
  261: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.  Both xen and
  262: the NetBSD kernel may be (and typically are) left compressed.
  263: 
  264: In a dom0, kernfs is mandatory for xend to communicate with the
  265: kernel, so ensure that /kern is in fstab.  (A standard NetBSD install
  266: should already mount /kern.)
  267: 
  268: Because you already installed NetBSD, you have a working boot setup
  269: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
  270: beginning of your root file system, have /boot, and likely also
  271: /boot.cfg.  (If not, fix before continuing!)
  272: 
  273: Add a line to /boot.cfg to boot Xen.  See boot.cfg(5) for an
  274: example.  The basic line is:
  275: 
  276: [[!template id=programlisting text="""
  277: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
  278: """]]
  279: 
  280: which specifies that the dom0 should have 512M, leaving the rest to be
  281: allocated for domUs.  To use a serial console, use
  282: 
  283: [[!template id=programlisting text="""
  284: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
  285: """]]
  286: 
  287: which will use the first serial port for Xen (which counts starting
  288: from 1, unlike NetBSD which counts starting from 0), forcing
  289: speed/parity.  Because the NetBSD command line lacks a
  290: "console=pc" argument, it will use the default "xencons" console device,
  291: which directs the console I/O through Xen to the same console device Xen
  292: itself uses (in this case, the serial port).
  293: 
  294: In an attempt to add performance, one can also add:
  295: 
  296: [[!template id=programlisting text="""
  297: dom0_max_vcpus=1 dom0_vcpus_pin
  298: """]]
  299: 
  300: to force only one vcpu to be provided (since NetBSD dom0 can't use
  301: more) and to pin that vcpu to a physical CPU.  TODO: benchmark this.
  302: 
  303: Xen has [many boot
  304: options](http://xenbits.xenproject.org/docs/4.5-testing/misc/xen-command-line.html),
  305: and other than dom0 memory and max_vcpus, they are generally not
  306: necessary.
  307: 
  308: As with non-Xen systems, you should have a line to boot /netbsd (a
  309: kernel that works without Xen).  Consider a line to boot /netbsd.ok (a
  310: fallback version of the non-Xen kernel, updated manually when you are
  311: sure /netbsd is ok).  Consider also a line to boot fallback versions
  312: of Xen and the dom0 kernel, but note that non-Xen NetBSD can be used
  313: to resolve Xen booting issues.
  314: 
  315: Now, reboot so that you are running a DOM0 kernel under Xen, rather
  316: than GENERIC without Xen.
  317: 
  318: Using grub (historic)
  319: ---------------------
  320: 
  321: Before NetBSD's native bootloader could support Xen, the use of
  322: grub was recommended.  If necessary, see the
  323: [old grub information](/ports/xen/howto-grub).
  324: 
  325: The [HowTo on Installing into
  326: RAID-1](https://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
  327: explains how to set up booting a dom0 with Xen using grub with
  328: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
  329: boot.  Now, just create a system with RAID-1, and alter /boot.cfg as
  330: described above.)
  331: 
  332: Configuring Xen
  333: ---------------
  334: 
  335: Now, you have a system that will boot Xen and the dom0 kernel, but not
  336: do anything else special.  Make sure that you have rebooted into Xen.
  337: There will be no domUs, and none can be started because you still have
  338: to configure the dom0 daemons.
  339: 
  340: The daemons which should be run vary with Xen version and with whether
  341: one is using xm or xl. Xen 4.2 and up packages use xl. To use xm with 4.2,
  342: edit xendomains to use xm instead.
  343: 
  344: For 4.1 and up, you should enable xencommons.  Not enabling xencommons
  345: will result in a hang; it is necessary to hit ^C on the console to let
  346: the machine finish booting.  If you are using xm (default in 4.1, or
  347: if you changed xendomains in 4.2), you should also enable xend:
  348: 
  349: [[!template id=programlisting text="""
  350: xend=YES # only if using xm, and only installed <= 4.2
  351: xencommons=YES
  352: """]]
  353: 
  354: TODO: Recommend for/against xen-watchdog.
  355: 
  356: After you have configured the daemons and either started them (in the
  357: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
  358: available resources, and running domains.  An example with xl follows:
  359: 
  360: [[!template id=programlisting text="""
  361: # xl dmesg
  362: ... xen's boot info ...
  363: # xl info
  364: ... available memory, etc ...
  365: # xl list
  366: Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  367: Domain-0           0       64    0  r----     58.1
  368: """]]
  369: 
  370: Xen logs will be in /var/log/xen.
  371: 
  372: ### Issues with xencommons
  373: 
  374: xencommons starts xenstored, which stores data on behalf of dom0 and
  375: domUs.  It does not currently work to stop and start xenstored.
  376: Certainly all domUs should be shutdown first, following the sort order
  377: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
  378: and is not notified when xenstored exits, leading to not recreating
  379: the state when the new xenstored starts.  Until there's a mechanism to
  380: make this work, one should not expect to be able to restart xenstored
  381: (and thus xencommons).  There is currently no reason to expect that
  382: this will get fixed any time soon.
  383: 
  384: ### No-longer needed advice about devices
  385: 
  386: The installation of NetBSD should already have created devices for xen
  387: (xencons, xenevt, xsd_kva), but if they are not present, create them:
  388: 
  389: [[!template id=programlisting text="""
  390: cd /dev && sh MAKEDEV xen
  391: """]]
  392: 
  393: anita (for testing NetBSD)
  394: --------------------------
  395: 
  396: With the setup so far (assuming 4.2/xl), one should be able to run
  397: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
  398: root, because anita must create a domU):
  399: 
  400: [[!template id=programlisting text="""
  401: anita --vmm=xl test file:///usr/obj/i386/
  402: """]]
  403: 
  404: Alternatively, one can use --vmm=xm to use xm-based domU creation
  405: instead (and must, on Xen <= 4.1).   TODO: confirm that anita xl really works.
  406:     
  407: Xen-specific NetBSD issues
  408: --------------------------
  409: 
  410: There are (at least) two additional things different about NetBSD as a
  411: dom0 kernel compared to hardware.
  412: 
  413: One is that the module ABI is different because some of the #defines
  414: change, so one must build modules for Xen.  As of netbsd-7, the build
  415: system does this automatically.  TODO: check this.  (Before building
  416: Xen modules was added, it was awkward to use modules to the point
  417: where it was considered that it did not work.)
  418: 
  419: The other difference is that XEN3_DOM0 does not have exactly the same
  420: options as GENERIC.  While it is debatable whether or not this is a
  421: bug, users should be aware of this and can simply add missing config
  422: items if desired.
  423: 
  424: Updating NetBSD in a dom0
  425: -------------------------
  426: 
  427: This is just like updating NetBSD on bare hardware, assuming the new
  428: version supports the version of Xen you are running.  Generally, one
  429: replaces the kernel and reboots, and then overlays userland binaries
  430: and adjusts /etc.
  431: 
  432: Note that one must update both the non-Xen kernel typically used for
  433: rescue purposes and the DOM0 kernel used with Xen.
  434: 
  435: Converting from grub to /boot
  436: -----------------------------
  437: 
  438: These instructions were [TODO: will be] used to convert a system from
  439: grub to /boot.  The system was originally installed in February of
  440: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
  441: over time.  Before these commands, it was running NetBSD 6 i386, Xen
  442: 4.1 and grub, much like the message linked earlier in the grub
  443: section.
  444: 
  445:         # Install MBR bootblocks on both disks. 
  446:         fdisk -i /dev/rwd0d
  447:         fdisk -i /dev/rwd1d
  448:         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
  449:         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
  450:         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
  451:         # Install secondary boot loader
  452:         cp -p /usr/mdec/boot /
  453:         # Create boot.cfg following earlier guidance:
  454:         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
  455:         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=512M
  456:         menu=GENERIC:boot
  457:         menu=GENERIC single-user:boot -s
  458:         menu=GENERIC.ok:boot netbsd.ok
  459:         menu=GENERIC.ok single-user:boot netbsd.ok -s
  460:         menu=Drop to boot prompt:prompt
  461:         default=1
  462:         timeout=30
  463: 
  464: TODO: actually do this and fix it if necessary.
  465: 
  466: Upgrading Xen versions
  467: ---------------------
  468: 
  469: Minor version upgrades are trivial.  Just rebuild/replace the
  470: xenkernel version and copy the new xen.gz to / (where /boot.cfg
  471: references it), and reboot.
  472: 
  473: Major version upgrades are conceptually not difficult, but can run
  474: into all the issues found when installing Xen.  Assuming migration
  475: from 4.1 to 4.2, remove the xenkernel41 and xentools41 packages and
  476: install the xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz
  477: to /.
  478: 
  479: Ensure that the contents of /etc/rc.d/xen* are correct.  Specifically,
  480: they must match the package you just installed and not be left over
  481: from some previous installation.
  482: 
  483: Enable the correct set of daemons; see the configuring section above.
  484: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
  485: 
  486: Ensure that the domU config files are valid for the new version.
  487: Specifically, for 4.x remove autorestart=True, and ensure that disks
  488: are specified with numbers as the second argument, as the examples
  489: above show, and not NetBSD device names.
  490: 
  491: Unprivileged domains (domU)
  492: ===========================
  493: 
  494: This section describes general concepts about domUs.  It does not
  495: address specific domU operating systems or how to install them.  The
  496: config files for domUs are typically in /usr/pkg/etc/xen, and are
  497: typically named so that the file name, domU name and the domU's host
  498: name match.
  499: 
  500: The domU is provided with CPU and memory by Xen, configured by the
  501: dom0.  The domU is provided with disk and network by the dom0,
  502: mediated by Xen, and configured in the dom0.
  503: 
  504: Entropy in domUs can be an issue; physical disks and network are on
  505: the dom0.  NetBSD's /dev/random system works, but is often challenged.
  506: 
  507: Config files
  508: ------------
  509: 
  510: There is no good order to present config files and the concepts
  511: surrounding what is being configured.  We first show an example config
  512: file, and then in the various sections give details.
  513: 
  514: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
  515: for a large number of well-commented examples, mostly for running
  516: GNU/Linux.
  517: 
  518: The following is an example minimal domain configuration file
  519: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
  520: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
  521: i386 domU).  The domU serves as a network file server.
  522: 
  523:         # -*- mode: python; -*-
  524: 
  525:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
  526:         memory = 1024
  527:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
  528:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
  529:                  'file:/n0/xen/foo-wd1,0x1,w' ]
  530: 
  531: The domain will have the same name as the file.  The kernel has the
  532: host/domU name in it, so that on the dom0 one can update the various
  533: domUs independently.  The vif line causes an interface to be provided,
  534: with a specific mac address (do not reuse MAC addresses!), in bridge
  535: mode.  Two disks are provided, and they are both writable; the bits
  536: are stored in files and Xen attaches them to a vnd(4) device in the
  537: dom0 on domain creation.  The system treats xbd0 as the boot device
  538: without needing explicit configuration.
  539: 
  540: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
  541: that "xm create" takes the name of a config file, while other commands
  542: take the name of a domain.  To create the domain, connect to the
  543: console, create the domain while attaching the console, shutdown the
  544: domain, and see if it has finished stopping, do (or xl with Xen >=
  545: 4.2):
  546: 
  547:         xm create foo
  548:         xm console foo
  549:         xm create -c foo
  550:         xm shutdown foo
  551:         xm list
  552: 
  553: Typing ^] will exit the console session.  Shutting down a domain is
  554: equivalent to pushing the power button; a NetBSD domU will receive a
  555: power-press event and do a clean shutdown.  Shutting down the dom0
  556: will trigger controlled shutdowns of all configured domUs.
  557: 
  558: domU kernels
  559: ------------
  560: 
  561: On a physical computer, the BIOS reads sector 0, and a chain of boot
  562: loaders finds and loads a kernel.  Normally this comes from the root
  563: file system.  With Xen domUs, the process is totally different.  The
  564: normal path is for the domU kernel to be a file in the dom0's
  565: file system.  At the request of the dom0, Xen loads that kernel into a
  566: new domU instance and starts execution.  While domU kernels can be
  567: anyplace, reasonable places to store domU kernels on the dom0 are in /
  568: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
  569: config files), or in /u0/xen (where the vdisks are).
  570: 
  571: Note that loading the domU kernel from the dom0 implies that boot
  572: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
  573: See the VPS section near the end for discussion of alternate ways to
  574: obtain domU kernels.
  575: 
  576: CPU and memory
  577: --------------
  578: 
  579: A domain is provided with some number of vcpus, less than the number
  580: of CPUs seen by the hypervisor.  (For a dom0, this is controlled by
  581: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
  582: from the config file by the "vcpus = N" directive.
  583: 
  584: A domain is provided with memory; this is controlled in the config
  585: file by "memory = N" (in megabytes).  In the straightforward case, the
  586: sum of the the memory allocated to the dom0 and all domUs must be less
  587: than the available memory.
  588: 
  589: Xen also provides a "balloon" driver, which can be used to let domains
  590: use more memory temporarily.  TODO: Explain better, and explain how
  591: well it works with NetBSD.
  592: 
  593: Virtual disks
  594: -------------
  595: 
  596: With the file/vnd style, typically one creates a directory,
  597: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
  598: domUs.  Then, for each domU disk, one writes zeros to a file that then
  599: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
  600: for the first virtual disk for the domU called foo.  Writing zeros to
  601: the file serves two purposes.  One is that preallocating the contents
  602: improves performance.  The other is that vnd on sparse files has
  603: failed to work.  TODO: give working/notworking NetBSD versions for
  604: sparse vnd and gnats reference.  Note that the use of file/vnd for Xen
  605: is not really different than creating a file-backed virtual disk for
  606: some other purpose, except that xentools handles the vnconfig
  607: commands.  To create an empty 4G virtual disk, simply do
  608: 
  609:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
  610: 
  611: Do not use qemu-img-xen, because this will create sparse file.  There
  612: have been recent (2015) reports of sparse vnd(4) devices causing
  613: lockups, but there is apparently no PR.
  614: 
  615: With the lvm style, one creates logical devices.  They are then used
  616: similarly to vnds.  TODO: Add an example with lvm.
  617: 
  618: In domU config files, the disks are defined as a sequence of 3-tuples.
  619: The first element is "method:/path/to/disk".  Common methods are
  620: "file:" for file-backed vnd. and "phy:" for something that is already
  621: a (TODO: character or block) device.
  622: 
  623: The second element is an artifact of how virtual disks are passed to
  624: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
  625: are given a device name to associate with the disk, and values like
  626: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
  627: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
  628: second argument.  The name given is converted to a major/minor by
  629: calling stat(2) on the name in /dev and this is passed to the domU.
  630: In the general case, the dom0 and domU can be different operating
  631: systems, and it is an unwarranted assumption that they have consistent
  632: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
  633: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
  634: for the second works fine and avoids this issue.  For a GNU/Linux
  635: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
  636: /dev/hda1.
  637: 
  638: The third element is "w" for writable disks, and "r" for read-only
  639: disks.
  640: 
  641: Note that NetBSD by default creates only vnd[0123].  If you need more
  642: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
  643: dom0.
  644: 
  645: Note that NetBSD by default creates only xbd[0123].  If you need more
  646: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
  647: 
  648: Virtual Networking
  649: ------------------
  650: 
  651: Xen provides virtual Ethernets, each of which connects the dom0 and a
  652: domU.  For each virtual network, there is an interface "xvifN.M" in
  653: the dom0, and in domU index N, a matching interface xennetM (NetBSD
  654: name).  The interfaces behave as if there is an Ethernet with two
  655: adapters connected.  From this primitive, one can construct various
  656: configurations.  We focus on two common and useful cases for which
  657: there are existing scripts: bridging and NAT.
  658: 
  659: With bridging (in the example above), the domU perceives itself to be
  660: on the same network as the dom0.  For server virtualization, this is
  661: usually best.  Bridging is accomplished by creating a bridge(4) device
  662: and adding the dom0's physical interface and the various xvifN.0
  663: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
  664: config file.  The bridge must be set up already in the dom0; an
  665: example /etc/ifconfig.bridge0 is:
  666: 
  667:         create
  668:         up
  669:         !brconfig bridge0 add wm0
  670: 
  671: With NAT, the domU perceives itself to be behind a NAT running on the
  672: dom0.  This is often appropriate when running Xen on a workstation.
  673: TODO: NAT appears to be configured by "vif = [ '' ]".
  674: 
  675: The MAC address specified is the one used for the interface in the new
  676: domain.  The interface in dom0 will use this address XOR'd with
  677: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
  678: 
  679: Sizing domains
  680: --------------
  681: 
  682: Modern x86 hardware has vast amounts of resources.  However, many
  683: virtual servers can function just fine on far less.  A system with
  684: 512M of RAM and a 4G disk can be a reasonable choice.  Note that it is
  685: far easier to adjust virtual resources than physical ones.  For
  686: memory, it's just a config file edit and a reboot.  For disk, one can
  687: create a new file and vnconfig it (or lvm), and then dump/restore,
  688: just like updating physical disks, but without having to be there and
  689: without those pesky connectors.
  690: 
  691: Starting domains automatically
  692: ------------------------------
  693: 
  694: To start domains foo at bar at boot and shut them down cleanly on dom0
  695: shutdown, in rc.conf add:
  696: 
  697:         xendomains="foo bar"
  698: 
  699: Note that earlier versions of the xentools41 xendomains rc.d script
  700: used xl, when one should use xm with 4.1.
  701: 
  702: Creating specific unprivileged domains (domU)
  703: =============================================
  704: 
  705: Creating domUs is almost entirely independent of operating system.  We
  706: have already presented the basics of config files.  Note that you must
  707: have already completed the dom0 setup so that "xl list" (or "xm list")
  708: works.
  709: 
  710: Creating an unprivileged NetBSD domain (domU)
  711: ---------------------------------------------
  712: 
  713: See the earlier config file, and adjust memory.  Decide on how much
  714: storage you will provide, and prepare it (file or lvm).
  715: 
  716: While the kernel will be obtained from the dom0 file system, the same
  717: file should be present in the domU as /netbsd so that tools like
  718: savecore(8) can work.   (This is helpful but not necessary.)
  719: 
  720: The kernel must be specifically for Xen and for use as a domU.  The
  721: i386 and amd64 provide the following kernels:
  722: 
  723:         i386 XEN3PAE_DOMU
  724:         amd64 XEN3_DOMU
  725: 
  726: This will boot NetBSD, but this is not that useful if the disk is
  727: empty.  One approach is to unpack sets onto the disk outside of xen
  728: (by mounting it, just as you would prepare a physical disk for a
  729: system you can't run the installer on).
  730: 
  731: A second approach is to run an INSTALL kernel, which has a miniroot
  732: and can load sets from the network.  To do this, copy the INSTALL
  733: kernel to / and change the kernel line in the config file to:
  734: 
  735:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  736: 
  737: Then, start the domain as "xl create -c configname".
  738: 
  739: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
  740: line should be used in the config file.
  741: 
  742:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  743: 
  744: After booting the domain, the option to install via CDROM may be
  745: selected.  The CDROM device should be changed to `xbd1d`.
  746: 
  747: Once done installing, "halt -p" the new domain (don't reboot or halt,
  748: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
  749: config file), switch the config file back to the XEN3_DOMU kernel,
  750: and start the new domain again. Now it should be able to use "root on
  751: xbd0a" and you should have a, functional NetBSD domU.
  752: 
  753: TODO: check if this is still accurate.
  754: When the new domain is booting you'll see some warnings about *wscons*
  755: and the pseudo-terminals. These can be fixed by editing the files
  756: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  757: `/etc/ttys`, except *console*, like this:
  758: 
  759:     console "/usr/libexec/getty Pc"         vt100   on secure
  760:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  761:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  762:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  763:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  764: 
  765: Finally, all screens must be commented out from `/etc/wscons.conf`.
  766: 
  767: It is also desirable to add
  768: 
  769:         powerd=YES
  770: 
  771: in rc.conf. This way, the domain will be properly shut down if
  772: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
  773: 
  774: It is not strictly necessary to have a kernel (as /netbsd) in the domU
  775: file system.  However, various programs (e.g. netstat) will use that
  776: kernel to look up symbols to read from kernel virtual memory.  If
  777: /netbsd is not the running kernel, those lookups will fail.  (This is
  778: not really a Xen-specific issue, but because the domU kernel is
  779: obtained from the dom0, it is far more likely to be out of sync or
  780: missing with Xen.)
  781: 
  782: Creating an unprivileged Linux domain (domU)
  783: --------------------------------------------
  784: 
  785: Creating unprivileged Linux domains isn't much different from
  786: unprivileged NetBSD domains, but there are some details to know.
  787: 
  788: First, the second parameter passed to the disk declaration (the '0x1' in
  789: the example below)
  790: 
  791:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  792: 
  793: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  794: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
  795: So, hda1 which has major 3 and minor 1 on a Linux system will have
  796: device number 0x301.  Alternatively, devices names can be used (hda,
  797: hdb, ...)  as xentools has a table to map these names to devices
  798: numbers.  To export a partition to a Linux guest we can use:
  799: 
  800:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
  801:         root = "/dev/hda1 ro"
  802: 
  803: and it will appear as /dev/hda on the Linux system, and be used as root
  804: partition.
  805: 
  806: To install the Linux system on the partition to be exported to the
  807: guest domain, the following method can be used: install
  808: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
  809: that will be the root partition of your Linux domain, and mount it.
  810: Then copy the files from a working Linux system, make adjustments in
  811: `/etc` (fstab, network config).  It should also be possible to extract
  812: binary packages such as .rpm or .deb directly to the mounted partition
  813: using the appropriate tool, possibly running under NetBSD's Linux
  814: emulation.  Once the file system has been populated, umount it.  If
  815: desirable, the file system can be converted to ext3 using tune2fs -j.
  816: It should now be possible to boot the Linux guest domain, using one of
  817: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
  818: 
  819: To get the Linux console right, you need to add:
  820: 
  821:     extra = "xencons=tty1"
  822: 
  823: to your configuration since not all Linux distributions auto-attach a
  824: tty to the xen console.
  825: 
  826: Creating an unprivileged Solaris domain (domU)
  827: ----------------------------------------------
  828: 
  829: See possibly outdated
  830: [Solaris domU instructions](/ports/xen/howto-solaris/).
  831: 
  832: 
  833: PCI passthrough: Using PCI devices in guest domains
  834: ---------------------------------------------------
  835: 
  836: The dom0 can give other domains access to selected PCI
  837: devices. This can allow, for example, a non-privileged domain to have
  838: access to a physical network interface or disk controller.  However,
  839: keep in mind that giving a domain access to a PCI device most likely
  840: will give the domain read/write access to the whole physical memory,
  841: as PCs don't have an IOMMU to restrict memory access to DMA-capable
  842: device.  Also, it's not possible to export ISA devices to non-dom0
  843: domains, which means that the primary VGA adapter can't be exported.
  844: A guest domain trying to access the VGA registers will panic.
  845: 
  846: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
  847: not been ported to later versions at this time.
  848: 
  849: For a PCI device to be exported to a domU, is has to be attached to
  850: the "pciback" driver in dom0.  Devices passed to the dom0 via the
  851: pciback.hide boot parameter will attach to "pciback" instead of the
  852: usual driver.  The list of devices is specified as "(bus:dev.func)",
  853: where bus and dev are 2-digit hexadecimal numbers, and func a
  854: single-digit number:
  855: 
  856:         pciback.hide=(00:0a.0)(00:06.0)
  857: 
  858: pciback devices should show up in the dom0's boot messages, and the
  859: devices should be listed in the `/kern/xen/pci` directory.
  860: 
  861: PCI devices to be exported to a domU are listed in the "pci" array of
  862: the domU's config file, with the format "0000:bus:dev.func".
  863: 
  864:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  865: 
  866: In the domU an "xpci" device will show up, to which one or more pci
  867: buses will attach.  Then the PCI drivers will attach to PCI buses as
  868: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
  869: or any PCI drivers built in by default; you have to build your own
  870: kernel to use PCI devices in a domU.  Here's a kernel config example;
  871: note that only the "xpci" lines are unusual.
  872: 
  873:         include         "arch/i386/conf/XEN3_DOMU"
  874: 
  875:         # Add support for PCI buses to the XEN3_DOMU kernel
  876:         xpci* at xenbus ?
  877:         pci* at xpci ?
  878: 
  879:         # PCI USB controllers
  880:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  881: 
  882:         # USB bus support
  883:         usb*    at uhci?
  884: 
  885:         # USB Hubs
  886:         uhub*   at usb?
  887:         uhub*   at uhub? port ? configuration ? interface ?
  888: 
  889:         # USB Mass Storage
  890:         umass*  at uhub? port ? configuration ? interface ?
  891:         wd*     at umass?
  892:         # SCSI controllers
  893:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  894: 
  895:         # SCSI bus support (for both ahc and umass)
  896:         scsibus* at scsi?
  897: 
  898:         # SCSI devices
  899:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
  900:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  901: 
  902: 
  903: NetBSD as a domU in a VPS
  904: =========================
  905: 
  906: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
  907: hardware.  This section explains how to deal with Xen in a domU as a
  908: virtual private server where you do not control or have access to the
  909: dom0.  This is not intended to be an exhaustive list of VPS providers;
  910: only a few are mentioned that specifically support NetBSD.
  911: 
  912: VPS operators provide varying degrees of access and mechanisms for
  913: configuration.  The big issue is usually how one controls which kernel
  914: is booted, because the kernel is nominally in the dom0 file system (to
  915: which VPS users do not normally have access).  A second issue is how
  916: to install NetBSD.
  917: A VPS user may want to compile a kernel for security updates, to run
  918: npf, run IPsec, or any other reason why someone would want to change
  919: their kernel.
  920: 
  921: One approach is to have an administrative interface to upload a kernel,
  922: or to select from a prepopulated list.  Other approaches are pygrub
  923: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
  924: kernel from the domU file system.  This is closer to a regular physical
  925: computer, where someone who controls a machine can replace the kernel.
  926: 
  927: A second issue is multiple CPUs.  With NetBSD 6, domUs support
  928: multiple vcpus, and it is typical for VPS providers to enable multiple
  929: CPUs for NetBSD domUs.
  930: 
  931: pygrub
  932: -------
  933: 
  934: pygrub runs in the dom0 and looks into the domU file system.  This
  935: implies that the domU must have a kernel in a file system in a format
  936: known to pygrub.  As of 2014, pygrub seems to be of mostly historical
  937: interest.
  938: 
  939: pvgrub
  940: ------
  941: 
  942: pvgrub is a version of grub that uses PV operations instead of BIOS
  943: calls.  It is booted from the dom0 as the domU kernel, and then reads
  944: /grub/menu.lst and loads a kernel from the domU file system.
  945: 
  946: [Panix](http://www.panix.com/) lets users use pvgrub.  Panix reports
  947: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
  948: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
  949: page](http://www.panix.com/v-colo/grub.html), which describes only
  950: Linux but should be updated to cover NetBSD :-).
  951: 
  952: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
  953: their own kernel.  See then [prgmr.com NetBSD
  954: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
  955: (which is in need of updating).
  956: 
  957: It appears that [grub's FFS
  958: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
  959: does not support all aspects of modern FFS, but there are also reports
  960: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
  961: partition for the kernel with the intent that grub can understand it,
  962: which leads to /netbsd not being the actual kernel.  One must remember
  963: to update the special boot partition.
  964: 
  965: Amazon
  966: ------
  967: 
  968: See the [Amazon EC2 page](/amazon_ec2/).
  969: 
  970: TODO items for improving NetBSD/xen
  971: ===================================
  972: 
  973: * Make the NetBSD dom0 kernel work with SMP.
  974: * Test the Xen 4.5 packages adequately to be able to recommend them as
  975:   the standard approach.
  976: * Get PCI passthrough working on Xen 4.5
  977: * Get pvgrub into pkgsrc, either via xentools or separately.
  978: * grub
  979:   * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
  980:     fragsize/blocksize (UFS2 support may be present; the point is to
  981:     make it so that with any UFS1/UFS2 file system setup that works
  982:     with NetBSD grub will also work).
  983:     See [pkg/40258](https://gnats.netbsd.org/40258).
  984:   * Push patches upstream.
  985:   * Get UFS2 patches into pvgrub.
  986: * Add support for PV ops to a version of /boot, and make it usable as
  987:   a kernel in Xen, similar to pvgrub.
  988: 
  989: Random pointers
  990: ===============
  991: 
  992: This section contains links from elsewhere not yet integrated into the
  993: HOWTO, and other guides.
  994: 
  995: * http://www.lumbercartel.ca/library/xen/
  996: * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html
  997: * https://gmplib.org/~tege/xen.html

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