File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.141: download - view: text, annotated - select for diffs
Fri Dec 15 16:40:59 2017 UTC (22 months ago) by khorben
Branches: MAIN
CVS tags: HEAD
No longer link to PR 47720

This PR (port-xen/47720) has been closed for over two years now.

    1: Introduction
    2: ============
    3: 
    4: [![[Xen
    5: screenshot]](https://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](https://www.netbsd.org/gallery/in-Action/hubertf-xen.png)
    6: 
    7: Xen is a hypervisor (or virtual machine monitor) for x86 hardware
    8: (i686-class or higher), which supports running multiple guest
    9: operating systems on a single physical machine.  Xen is a Type 1 or
   10: bare-metal hypervisor; one uses the Xen kernel to control the CPU,
   11: memory and console, a dom0 operating system which mediates access to
   12: other hardware (e.g., disks, network, USB), and one or more domU
   13: operating systems which operate in an unprivileged virtualized
   14: environment.  IO requests from the domU systems are forwarded by the
   15: hypervisor (Xen) to the dom0 to be 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 machine that runs NetBSD/amd64 will work with Xen and PV
   31: guests.  In theory i386 computers (without x86_64/amd64 support) can
   32: be used for Xen <= 4.2, but we have no recent reports of this working
   33: (this is a hint).  For HVM guests, hardware support is needed, but it
   34: is common on recent machines.  For Intel CPUs, one needs the VT-x
   35: extension, shown in "cpuctl identify 0" as VMX.  For AMD CPUs, one
   36: needs the AMD-V extensions, shown in "cpuctl identify 0" as SVM.
   37: There are further features for IOMMU virtualization, Intel's VT-d and
   38: AMD's AMD-Vi.  TODO: Explain whether Xen on NetBSD makes use of these
   39: features.  TODO: Review by someone who really understands this.
   40: 
   41: Note that a FreeBSD dom0 requires VT-x and VT-d (or equivalent); this
   42: is because the FreeBSD dom0 does not run in PV mode.
   43: 
   44: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
   45: The dom0 can start one or more domUs.  (Booting is explained in detail
   46: in the dom0 section.)
   47: 
   48: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
   49: and that Xen kernels and tools are available in pkgsrc.  This HOWTO
   50: attempts to address both the case of running a NetBSD dom0 on hardware
   51: and running domUs under it (NetBSD and other), and also running NetBSD
   52: as a domU in a VPS.
   53: 
   54: Xen 3.1 in pkgsrc supports "PCI passthrough", which means that
   55: specific PCI devices can be made available to a specific domU instead
   56: of the dom0.  This can be useful to let a domU run X11, or access some
   57: network interface or other peripheral.
   58: 
   59: NetBSD 6 and earlier supported Xen 2; support was removed from NetBSD
   60: 7.  Xen 2 has been removed from pkgsrc.
   61: 
   62: Prerequisites
   63: -------------
   64: 
   65: Installing NetBSD/Xen is not extremely difficult, but it is more
   66: complex than a normal installation of NetBSD.
   67: In general, this HOWTO is occasionally overly restrictive about how
   68: things must be done, guiding the reader to stay on the established
   69: path when there are no known good reasons to stray.
   70: 
   71: This HOWTO presumes a basic familiarity with the Xen system
   72: architecture, with installing NetBSD on i386/amd64 hardware, and with
   73: installing software from pkgsrc.  See also the [Xen
   74: website](http://www.xenproject.org/).
   75: 
   76: Versions of Xen and NetBSD
   77: ==========================
   78: 
   79: Most of the installation concepts and instructions are independent
   80: of Xen version and NetBSD version.  This section gives advice on
   81: which version to choose.  Versions not in pkgsrc and older unsupported
   82: versions of NetBSD are intentionally ignored.
   83: 
   84: The term "amd64" is used to refer to both the NetBSD port and to the
   85: hardware architecture on which it runs.  (Such hardware is made by
   86: both Intel and AMD, and in 2016 a normal PC has this CPU
   87: architecture.)
   88: 
   89: Xen
   90: ---
   91: 
   92: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
   93: xenkernel and xentools.  We will refer only to the kernel versions,
   94: but note that both packages must be installed together and must have
   95: matching versions.
   96: 
   97: xenkernel3 provides Xen 3.1.  It is no longer maintained by Xen, and
   98: the last applied security patch was in 2011. Thus, it should not be
   99: used.  It supports PCI passthrough, which is why people use it anyway.
  100: Xen 3.1 runs on i386 (both non-PAE and PAE) and amd64 hardware.
  101: 
  102: xenkernel33 provides Xen 3.3.  It is no longer maintained by Xen, and
  103: the last applied security patch was in 2012.  Thus, it should not be
  104: used.  Xen 3.3 runs on i386 PAE and amd64 hardware.  There are no good
  105: reasons to run this version.
  106: 
  107: xenkernel41 provides Xen 4.1.  It is no longer maintained by Xen, but
  108: as of 2016-12 received backported security patches.  Xen 4.1 runs on
  109: i386 PAE and amd64 hardware.  There are no good reasons to run this
  110: version.
  111: 
  112: Note that 3.1, 3.3 and 4.1 have been removed from pkgsrc-current, but
  113: are in 2016Q4.  They will be removed from this HOWTO sometime after
  114: 2017Q1.
  115: 
  116: xenkernel42 provides Xen 4.2.  It is no longer maintained by Xen, but
  117: as of 2016-12 received backported security patches.  Xen 4.2 runs on
  118: i386 PAE and amd64 hardware.  The only reason to run this is if you
  119: need to use xm instead of xl, or if you need to run on hardware that
  120: supports i386 but not amd64.  (This might also be useful if you need
  121: an i386 dom0, if it turns out that an amd64 Xen kernel and an i386
  122: dom0 is problematic.)
  123: 
  124: xenkernel45 provides Xen 4.5.  As of 2016-12, security patches were
  125: released by Xen and applied to pkgsrc.  Xen 4.5 runs on amd64 hardware
  126: only.  While slightly old, 4.5 has been tested and run by others, so
  127: it is the conservative choice.
  128: 
  129: xenkernel46 provides Xen 4.6.  It is new to pkgsrc as of 2016-05.  As
  130: of 2016-12, security patches were released by Xen and applied to
  131: pkgsrc.  Xen 4.6 runs on amd64 hardware only For new installations,
  132: 4.6 is probably the appropriate choice and it will likely soon be the
  133: standard approach.  (If using Ubuntu guests, be sure to have the
  134: xentools46 from December, 2016).
  135: 
  136: Xen 4.7 (released 2016-06) and 4.8 (released 2016-12) are not yet in
  137: pkgsrc.
  138: 
  139: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
  140: 
  141: Note that NetBSD support is called XEN3.  It works with Xen 3 and Xen
  142: 4 because the hypercall interface has been stable.
  143: 
  144: Xen command program
  145: -------------------
  146: 
  147: Early Xen used a program called xm to manipulate the system from the
  148: dom0.  Starting in 4.1, a replacement program with similar behavior
  149: called xl is provided, but it does not work well in 4.1.  In 4.2, both
  150: xm and xl work fine.  4.4 is the last version that has xm.
  151: 
  152: You must make a global choice to use xm or xl, because it affects not
  153: only which command you use, but the command used by rc.d scripts
  154: (specifically xendomains) and which daemons should be run.  The
  155: xentools packages provide xm for 3.1, 3.3 and 4.1 and xl for 4.2 and up.
  156: 
  157: In 4.2, you can choose to use xm by simply changing the ctl_command
  158: variable and setting xend=YES in rc.conf.
  159: 
  160: With xl, virtual devices are configured in parallel, which can cause
  161: problems if they are written assuming serial operation (e.g., updating
  162: firewall rules without explicit locking).  There is now locking for
  163: the provided scripts, which works for normal casses (e.g, file-backed
  164: xbd, where a vnd must be allocated).  But, as of 201612, it has not
  165: been adequately tested for a complex custom setup with a large number
  166: of interfaces.
  167: 
  168: NetBSD
  169: ------
  170: 
  171: The netbsd-6, netbsd-7, and -current branches are all reasonable
  172: choices, with more or less the same considerations for non-Xen use.
  173: Therefore, netbsd-7 is recommended as the stable version of the most
  174: recent release for production use.  In addition, netbsd-7 and -current
  175: have a important scheduler fix (in November of 2015) affecting
  176: contention between dom0 and domUs; see
  177: https://releng.netbsd.org/cgi-bin/req-7.cgi?show=1040 for a
  178: description.  For those wanting to learn Xen or without production
  179: stability concerns, netbsd-7 is still likely most appropriate, but
  180: -current is also a reasonable choice.  (Xen runs ok on netbsd-5, but
  181: the xentools packages are likely difficult to build, and netbsd-5 is
  182: not supported.)
  183: 
  184: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
  185: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
  186: need SMP for dom0 functions; the lack of support is really a problem
  187: when using a dom0 as a normal computer.)
  188: 
  189: Architecture
  190: ------------
  191: 
  192: Xen itself can run on i386 (Xen < 4.2) or amd64 hardware (all Xen
  193: versions).  (Practically, almost any computer where one would want to
  194: run Xen today supports amd64.)
  195: 
  196: Xen, the dom0 system, and each domU system can be either i386 or
  197: amd64.  When building a xenkernel package, one obtains an i386 Xen
  198: kernel on an i386 host, and an amd64 Xen kernel on an amd64 host.  If
  199: the Xen kernel is i386, then the dom0 kernel and all domU kernels must
  200: be i386.  With an amd64 Xen kernel, an amd64 dom0 kernel is known to
  201: work, and an i386 dom0 kernel should in theory work.  An amd64
  202: Xen/dom0 is known to support both i386 and amd64 domUs.
  203: 
  204: i386 dom0 and domU kernels must be PAE (except for an i386 Xen 3.1
  205: kernel, where one can use non-PAE for dom0 and all domUs); PAE kernels
  206: are included in the NetBSD default build.  (Note that emacs (at least)
  207: fails if run on i386 with PAE when built without, and vice versa,
  208: presumably due to bugs in the undump code.)
  209: 
  210: Because of the above, the standard approach is to use an amd64 Xen
  211: kernel and NetBSD/amd64 for the dom0.  For domUs, NetBSD/i386 (with
  212: the PAE kernel) and NetBSD/amd64 are in widespread use, and there is
  213: little to no Xen-specific reason to prefer one over the other.
  214: 
  215: Note that to use an i386 dom0 with Xen 4.5 or higher, one must build
  216: (or obtain from pre-built packages) an amd64 Xen kernel and install
  217: that on the system.  (One must also use a PAE i386 kernel, but this is
  218: also required with an i386 Xen kernel.).  Almost no one in the
  219: NetBSD/Xen community does this, and the standard, well-tested,
  220: approach is to use an amd64 dom0.
  221: 
  222: A [posting on
  223: xen-devel](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html)
  224: explained that PV system call overhead was higher on amd64, and thus
  225: there is some notion that i386 guests are faster.  It goes on to
  226: caution that the total situation is complex and not entirely
  227: understood. On top of that caution, the post is about Linux, not
  228: NetBSD.  TODO: Include link to benchmarks, if someone posts them.
  229: 
  230: Stability
  231: ---------
  232: 
  233: Mostly, NetBSD as a dom0 or domU is quite stable.
  234: However, there are some open PRs indicating problems.
  235: 
  236:  - [PR 48125](http://gnats.netbsd.org/48125)
  237: 
  238: Note also that there are issues with sparse vnd(4) instances, but
  239: these are not about Xen -- they just are noticed with sparse vnd(4)
  240: instances in support of virtual disks in a dom0.
  241: 
  242: Recommendation
  243: --------------
  244: 
  245: Therefore, this HOWTO recommends running xenkernel45 or xenkernel46,
  246: xl, the NetBSD 7 stable branch, and to use an amd64 kernel as the
  247: dom0.  Either the i386PAE or amd64 version of NetBSD may be used as
  248: domUs.
  249: 
  250: Because bugs are fixed quite often, and because of Xen security
  251: advisories, it is good to stay up to date with NetBSD (tracking a
  252: stable branch), with the Xen kernel (tracking a Xen version via
  253: pkgsrc), and with the Xen tools.  Specifically, NetBSD (-7 and
  254: -current) got an important fix affecting dom0/domU timesharing in
  255: November, 2015, and xentools46 got a fix to enable Ubuntu guests to
  256: boot in December, 2016.
  257: 
  258: Status
  259: ------
  260: 
  261: Ideally, all versions of Xen in pkgsrc would build on all supported
  262: versions of NetBSD/amd64, to the point where this section would be
  263: silly.  However, that has not always been the case.  Besides aging
  264: code and aging compilers, qemu (included in xentools for HVM support)
  265: is difficult to build.  Note that there is intentionally no data for
  266: 4.5+ up for i386, and often omits xentools info if the corresponding
  267: kernel fails.
  268: 
  269: The following table gives status, with the date last checked
  270: (generally on the most recent quarterly branch).  The first code is
  271: "builds" if it builds ok, and "FAIL" for a failure to build.  The
  272: second code/date only appears for xenkernel* and is "works" if it runs
  273: ok as a dom0 and can support a domU, and "FAIL" if it won't boot or
  274: run a domU.
  275: 
  276: 	xenkernel3 netbsd-6 i386 FAIL 201612
  277: 	xenkernel33 netbsd-6 i386 FAIL 201612
  278: 	xenkernel41 netbsd-6 i386 builds 201612
  279: 	xenkernel42 netbsd-6 i386 builds 201612
  280: 	xentools3 netbsd-6 i386 FAIL 201612
  281: 	xentools33 netbsd-6 i386 FAIL 201612
  282: 	xentools41 netbsd-6 i386 builds 201612
  283: 	xentools42 netbsd-6 i386 FAIL 201612
  284: 
  285: 	xenkernel3 netbsd-7 i386 FAIL 201412
  286: 	xenkernel33 netbsd-7 i386 FAIL 201412
  287: 	xenkernel41 netbsd-7 i386 builds 201412
  288: 	xenkernel42 netbsd-7 i386 builds 201412
  289: 	xentools41 netbsd-7 i386 builds 201412
  290: 	xentools42 netbsd-7 i386 ??FAIL 201412
  291: 
  292: 	xenkernel3 netbsd-6 amd64 FAIL 201612
  293: 	xenkernel33 netbsd-6 amd64 FAIL 201612
  294: 	xenkernel41 netbsd-6 amd64 builds 201612 works 201612
  295: 	xenkernel42 netbsd-6 amd64 builds 201612 works 201612
  296: 	xenkernel45 netbsd-6 amd64 builds 201612
  297: 	xenkernel46 netbsd-6 amd64 builds 201612
  298: 	xentools41 netbsd-6 amd64 builds 201612
  299: 	xentools42 netbsd-6 amd64 builds 201612
  300: 	xentools45 netbsd-6 amd64 builds 201612
  301: 	xentools46 netbsd-6 amd64 FAIL 201612
  302: 
  303: 	xenkernel3 netbsd-7 amd64 builds 201612
  304: 	xenkernel33 netbsd-7 amd64 builds 201612
  305: 	xenkernel41 netbsd-7 amd64 builds 201612
  306: 	xenkernel42 netbsd-7 amd64 builds 201612
  307: 	xenkernel45 netbsd-7 amd64 builds 201612
  308: 	xenkernel46 netbsd-7 amd64 builds 201612
  309: 	xentools3 netbsd-7 amd64 builds 201612
  310: 	xentools3-hvm netbsd-7 amd64 builds 201612
  311: 	xentools33 netbsd-7 amd64 FAIL 201612
  312: 	xentools41 netbsd-7 amd64 builds 201612
  313: 	xentools42 netbsd-7 amd64 builds 201612
  314: 	xentools45 netbsd-7 amd64 builds 201612
  315: 	xentools46 netbsd-7 amd64 builds 201612
  316: 
  317: NetBSD as a dom0
  318: ================
  319: 
  320: NetBSD can be used as a dom0 and works very well.  The following
  321: sections address installation, updating NetBSD, and updating Xen.
  322: Note that it doesn't make sense to talk about installing a dom0 OS
  323: without also installing Xen itself.  We first address installing
  324: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
  325: NetBSD install to a dom0 install by just changing the kernel and boot
  326: configuration.
  327: 
  328: For experimenting with Xen, a machine with as little as 1G of RAM and
  329: 100G of disk can work.  For running many domUs in productions, far
  330: more will be needed; e.g. 4-8G and 1T of disk is reasonable for a
  331: half-dozen domUs of 512M and 32G each.  Basically, the RAM and disk
  332: have to be bigger than the sum of the RAM/disk needs of the dom0 and
  333: all the domUs.
  334: 
  335: Styles of dom0 operation
  336: ------------------------
  337: 
  338: There are two basic ways to use Xen.  The traditional method is for
  339: the dom0 to do absolutely nothing other than providing support to some
  340: number of domUs.  Such a system was probably installed for the sole
  341: purpose of hosting domUs, and sits in a server room on a UPS.
  342: 
  343: The other way is to put Xen under a normal-usage computer, so that the
  344: dom0 is what the computer would have been without Xen, perhaps a
  345: desktop or laptop.  Then, one can run domUs at will.  Purists will
  346: deride this as less secure than the previous approach, and for a
  347: computer whose purpose is to run domUs, they are right.  But Xen and a
  348: dom0 (without domUs) is not meaningfully less secure than the same
  349: things running without Xen.  One can boot Xen or boot regular NetBSD
  350: alternately with little problems, simply refraining from starting the
  351: Xen daemons when not running Xen.
  352: 
  353: Note that NetBSD as dom0 does not support multiple CPUs.  This will
  354: limit the performance of the Xen/dom0 workstation approach.  In theory
  355: the only issue is that the "backend drivers" are not yet MPSAFE:
  356:   https://mail-index.netbsd.org/netbsd-users/2014/08/29/msg015195.html
  357: 
  358: Installation of NetBSD
  359: ----------------------
  360: 
  361: First,
  362: [install NetBSD/amd64](/guide/inst/)
  363: just as you would if you were not using Xen.
  364: However, the partitioning approach is very important.
  365: 
  366: If you want to use RAIDframe for the dom0, there are no special issues
  367: for Xen.  Typically one provides RAID storage for the dom0, and the
  368: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
  369: over a RAID1 header to find /boot from a file system within a RAID
  370: partition; this is no different when booting Xen.
  371: 
  372: There are 4 styles of providing backing storage for the virtual disks
  373: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
  374: 
  375: With raw partitions, one has a disklabel (or gpt) partition sized for
  376: each virtual disk to be used by the domU.  (If you are able to predict
  377: how domU usage will evolve, please add an explanation to the HOWTO.
  378: Seriously, needs tend to change over time.)
  379: 
  380: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
  381: for domU disks.  This is almost as efficient as raw disk partitions
  382: and more flexible.  Hence raw disk partitions should typically not
  383: be used.
  384: 
  385: One can use files in the dom0 file system, typically created by dd'ing
  386: /dev/zero to create a specific size.  This is somewhat less efficient,
  387: but very convenient, as one can cp the files for backup, or move them
  388: between dom0 hosts.
  389: 
  390: Finally, in theory one can place the files backing the domU disks in a
  391: SAN.  (This is an invitation for someone who has done this to add a
  392: HOWTO page.)
  393: 
  394: Installation of Xen
  395: -------------------
  396: 
  397: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
  398: pkgsrc (or another matching pair).  See [the pkgsrc
  399: documentation](https://www.NetBSD.org/docs/pkgsrc/) for help with
  400: pkgsrc.  Ensure that your packages are recent; the HOWTO does not
  401: contemplate old builds.
  402: 
  403: 
  404: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
  405: recent versions have HVM support integrated in the main xentools
  406: package.  It is entirely reasonable to run only PV guests.
  407: 
  408: Next you need to install the selected Xen kernel itself, which is
  409: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
  410: For debugging, one may copy xen-debug.gz; this is conceptually similar
  411: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
  412: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
  413: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
  414: of a NetBSD build.  If using i386, use
  415: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.  (If using Xen
  416: 3.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen.  But you
  417: should not use Xen 3.1.)  Both xen and the NetBSD kernel may be (and
  418: typically are) left compressed.
  419: 
  420: In a dom0, kernfs is mandatory for xend to communicate with the
  421: kernel, so ensure that /kern is in fstab.  (A standard NetBSD install
  422: should already mount /kern.)
  423: 
  424: Because you already installed NetBSD, you have a working boot setup
  425: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
  426: beginning of your root file system, have /boot, and likely also
  427: /boot.cfg.  (If not, fix before continuing!)
  428: 
  429: Add a line to to /boot.cfg to boot Xen.  See boot.cfg(5) for an
  430: example.  The basic line is
  431: 
  432:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
  433: 
  434: which specifies that the dom0 should have 256M, leaving the rest to be
  435: allocated for domUs.  To use a serial console, use
  436: 
  437:         menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=256M console=com1 com1=9600,8n1
  438: 
  439: which will use the first serial port for Xen (which counts starting
  440: from 1, unlike NetBSD which counts starting from 0), forcing
  441: speed/parity.  Because the NetBSD command line lacks a
  442: "console=pc" argument, it will use the default "xencons" console device,
  443: which directs the console I/O through Xen to the same console device Xen
  444: itself uses (in this case, the serial port).
  445: 
  446: In an attempt to add performance, one can also add
  447: 
  448:         dom0_max_vcpus=1 dom0_vcpus_pin
  449: 
  450: to force only one vcpu to be provided (since NetBSD dom0 can't use
  451: more) and to pin that vcpu to a physical CPU.  TODO: benchmark this.
  452: 
  453: Xen has [many boot
  454: options](http://xenbits.xenproject.org/docs/4.5-testing/misc/xen-command-line.html),
  455: and other than dom0 memory and max_vcpus, they are generally not
  456: necessary.
  457: 
  458: As with non-Xen systems, you should have a line to boot /netbsd (a
  459: kernel that works without Xen).  Consider a line to boot /netbsd.ok (a
  460: fallback version of the non-Xen kernel, updated manually when you are
  461: sure /netbsd is ok).  Consider also a line to boot fallback versions
  462: of Xen and the dom0 kernel, but note that non-Xen NetBSD can be used
  463: to resolve Xen booting issues.
  464: 
  465: Probably you want a default=N line to choose Xen in the absence of
  466: intervention.
  467: 
  468: Now, reboot so that you are running a DOM0 kernel under Xen, rather
  469: than GENERIC without Xen.
  470: 
  471: Using grub (historic)
  472: ---------------------
  473: 
  474: Before NetBSD's native bootloader could support Xen, the use of
  475: grub was recommended.  If necessary, see the
  476: [old grub information](/ports/xen/howto-grub).
  477: 
  478: The [HowTo on Installing into
  479: RAID-1](https://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
  480: explains how to set up booting a dom0 with Xen using grub with
  481: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
  482: boot.  Now, just create a system with RAID-1, and alter /boot.cfg as
  483: described above.)
  484: 
  485: Configuring Xen
  486: ---------------
  487: 
  488: Xen logs will be in /var/log/xen.
  489: 
  490: Now, you have a system that will boot Xen and the dom0 kernel, but not
  491: do anything else special.  Make sure that you have rebooted into Xen.
  492: There will be no domUs, and none can be started because you still have
  493: to configure the dom0 daemons.
  494: 
  495: The daemons which should be run vary with Xen version and with whether
  496: one is using xm or xl.  The Xen 3.1, 3.3 and 4.1 packages use xm.  Xen
  497: 4.2 and up packages use xl.  To use xm with 4.2, edit xendomains to
  498: use xm instead.
  499: 
  500: For 3.1 and 3.3, you should enable xend and xenbackendd:
  501: 
  502:         xend=YES
  503:         xenbackendd=YES
  504: 
  505: For 4.1 and up, you should enable xencommons.  Not enabling xencommons
  506: will result in a hang; it is necessary to hit ^C on the console to let
  507: the machine finish booting.  If you are using xm (default in 4.1, or
  508: if you changed xendomains in 4.2), you should also enable xend:
  509: 
  510:         xend=YES # only if using xm, and only installed <= 4.2
  511:         xencommons=YES
  512: 
  513: TODO: Recommend for/against xen-watchdog.
  514: 
  515: After you have configured the daemons and either started them (in the
  516: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
  517: available resources, and running domains.  An example with xl follows:
  518: 
  519:         # xl dmesg
  520: 	[xen's boot info]
  521:         # xl info
  522: 	[available memory, etc.]
  523:         # xl list
  524:         Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  525:         Domain-0           0       64    0  r----     58.1
  526: 
  527: ### Issues with xencommons
  528: 
  529: xencommons starts xenstored, which stores data on behalf of dom0 and
  530: domUs.  It does not currently work to stop and start xenstored.
  531: Certainly all domUs should be shutdown first, following the sort order
  532: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
  533: and is not notified when xenstored exits, leading to not recreating
  534: the state when the new xenstored starts.  Until there's a mechanism to
  535: make this work, one should not expect to be able to restart xenstored
  536: (and thus xencommons).  There is currently no reason to expect that
  537: this will get fixed any time soon.
  538: 
  539: ### No-longer needed advice about devices
  540: 
  541: The installation of NetBSD should already have created devices for xen
  542: (xencons, xenevt, xsd_kva), but if they are not present, create them:
  543: 
  544:         cd /dev && sh MAKEDEV xen
  545: 
  546: anita (for testing NetBSD)
  547: --------------------------
  548: 
  549: With the setup so far (assuming 4.2/xl), one should be able to run
  550: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
  551: root, because anita must create a domU):
  552: 
  553:         anita --vmm=xl test file:///usr/obj/i386/
  554: 
  555: Alternatively, one can use --vmm=xm to use xm-based domU creation
  556: instead (and must, on Xen <= 4.1).   TODO: confirm that anita xl really works.
  557:     
  558: Xen-specific NetBSD issues
  559: --------------------------
  560: 
  561: There are (at least) two additional things different about NetBSD as a
  562: dom0 kernel compared to hardware.
  563: 
  564: One is that the module ABI is different because some of the #defines
  565: change, so one must build modules for Xen.  As of netbsd-7, the build
  566: system does this automatically.  TODO: check this.  (Before building
  567: Xen modules was added, it was awkward to use modules to the point
  568: where it was considered that it did not work.)
  569: 
  570: The other difference is that XEN3_DOM0 does not have exactly the same
  571: options as GENERIC.  While it is debatable whether or not this is a
  572: bug, users should be aware of this and can simply add missing config
  573: items if desired.
  574: 
  575: Updating NetBSD in a dom0
  576: -------------------------
  577: 
  578: This is just like updating NetBSD on bare hardware, assuming the new
  579: version supports the version of Xen you are running.  Generally, one
  580: replaces the kernel and reboots, and then overlays userland binaries
  581: and adjusts /etc.
  582: 
  583: Note that one must update both the non-Xen kernel typically used for
  584: rescue purposes and the DOM0 kernel used with Xen.
  585: 
  586: Converting from grub to /boot
  587: -----------------------------
  588: 
  589: These instructions were [TODO: will be] used to convert a system from
  590: grub to /boot.  The system was originally installed in February of
  591: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
  592: over time.  Before these commands, it was running NetBSD 6 i386, Xen
  593: 4.1 and grub, much like the message linked earlier in the grub
  594: section.
  595: 
  596:         # Install MBR bootblocks on both disks. 
  597:         fdisk -i /dev/rwd0d
  598:         fdisk -i /dev/rwd1d
  599:         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
  600:         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
  601:         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
  602:         # Install secondary boot loader
  603:         cp -p /usr/mdec/boot /
  604:         # Create boot.cfg following earlier guidance:
  605:         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
  606:         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
  607:         menu=GENERIC:boot
  608:         menu=GENERIC single-user:boot -s
  609:         menu=GENERIC.ok:boot netbsd.ok
  610:         menu=GENERIC.ok single-user:boot netbsd.ok -s
  611:         menu=Drop to boot prompt:prompt
  612:         default=1
  613:         timeout=30
  614: 
  615: TODO: actually do this and fix it if necessary.
  616: 
  617: Upgrading Xen versions
  618: ---------------------
  619: 
  620: Minor version upgrades are trivial.  Just rebuild/replace the
  621: xenkernel version and copy the new xen.gz to / (where /boot.cfg
  622: references it), and reboot.
  623: 
  624: Major version upgrades are conceptually not difficult, but can run
  625: into all the issues found when installing Xen.  Assuming migration
  626: from 4.1 to 4.2, remove the xenkernel41 and xentools41 packages and
  627: install the xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz
  628: to /.
  629: 
  630: Ensure that the contents of /etc/rc.d/xen* are correct.  Specifically,
  631: they must match the package you just installed and not be left over
  632: from some previous installation.
  633: 
  634: Enable the correct set of daemons; see the configuring section above.
  635: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
  636: 
  637: Ensure that the domU config files are valid for the new version.
  638: Specifically, for 4.x remove autorestart=True, and ensure that disks
  639: are specified with numbers as the second argument, as the examples
  640: above show, and not NetBSD device names.
  641: 
  642: Hardware known to work
  643: ----------------------
  644: 
  645: Arguably, this section is misplaced, and there should be a page of
  646: hardware that runs NetBSD/amd64 well, with the mostly-well-founded
  647: assumption that NetBSD/xen runs fine on any modern hardware that
  648: NetBSD/amd64 runs well on.  Until then, we give motherboard/CPU (and
  649: sometimes RAM) pairs/triples to aid those choosing a motherboard.
  650: Note that Xen systems usually do not run X, so a listing here does not
  651: imply that X works at all.
  652: 
  653:         Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
  654:         Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
  655:         ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
  656: 
  657: Older hardware:
  658: 
  659:         Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
  660:         INTEL DG33FB, "Intel(R) Core(TM)2 Duo CPU     E6850  @ 3.00GHz"
  661:         INTEL DG33FB, "Intel(R) Core(TM)2 Duo CPU     E8400  @ 3.00GHz"
  662: 
  663: Running Xen under qemu
  664: ----------------------
  665: 
  666: The astute reader will note that this section is somewhat twisted.
  667: However, it can be useful to run Xen under qemu either because the
  668: version of NetBSD as a dom0 does not run on the hardware in use, or to
  669: generate automated test cases involving Xen.
  670: 
  671: In 2015-01, the following combination was reported to mostly work:
  672: 
  673:         host OS: NetBSD/amd64 6.1.4
  674:         qemu: 2.2.0 from pkgsrc
  675:         Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
  676:         dom0 kernel: NetBSD/amd64 6.1.5
  677:         Xen tools: xentools42-4.2.5 from pkgsrc
  678: 
  679: See [PR 47720](https://gnats.netbsd.org/47720) for a problem with dom0
  680: shutdown.
  681: 
  682: Unprivileged domains (domU)
  683: ===========================
  684: 
  685: This section describes general concepts about domUs.  It does not
  686: address specific domU operating systems or how to install them.  The
  687: config files for domUs are typically in /usr/pkg/etc/xen, and are
  688: typically named so that the file name, domU name and the domU's host
  689: name match.
  690: 
  691: The domU is provided with CPU and memory by Xen, configured by the
  692: dom0.  The domU is provided with disk and network by the dom0,
  693: mediated by Xen, and configured in the dom0.
  694: 
  695: Entropy in domUs can be an issue; physical disks and network are on
  696: the dom0.  NetBSD's /dev/random system works, but is often challenged.
  697: 
  698: Config files
  699: ------------
  700: 
  701: There is no good order to present config files and the concepts
  702: surrounding what is being configured.  We first show an example config
  703: file, and then in the various sections give details.
  704: 
  705: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
  706: for a large number of well-commented examples, mostly for running
  707: GNU/Linux.
  708: 
  709: The following is an example minimal domain configuration file
  710: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
  711: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
  712: i386 domU).  The domU serves as a network file server.
  713: 
  714:         # -*- mode: python; -*-
  715: 
  716:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
  717:         memory = 1024
  718:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
  719:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
  720:                  'file:/n0/xen/foo-wd1,0x1,w' ]
  721: 
  722: The domain will have the same name as the file.  The kernel has the
  723: host/domU name in it, so that on the dom0 one can update the various
  724: domUs independently.  The vif line causes an interface to be provided,
  725: with a specific mac address (do not reuse MAC addresses!), in bridge
  726: mode.  Two disks are provided, and they are both writable; the bits
  727: are stored in files and Xen attaches them to a vnd(4) device in the
  728: dom0 on domain creation.  The system treats xbd0 as the boot device
  729: without needing explicit configuration.
  730: 
  731: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
  732: that "xm create" takes the name of a config file, while other commands
  733: take the name of a domain.  To create the domain, connect to the
  734: console, create the domain while attaching the console, shutdown the
  735: domain, and see if it has finished stopping, do (or xl with Xen >=
  736: 4.2):
  737: 
  738:         xm create foo
  739:         xm console foo
  740:         xm create -c foo
  741:         xm shutdown foo
  742:         xm list
  743: 
  744: Typing ^] will exit the console session.  Shutting down a domain is
  745: equivalent to pushing the power button; a NetBSD domU will receive a
  746: power-press event and do a clean shutdown.  Shutting down the dom0
  747: will trigger controlled shutdowns of all configured domUs.
  748: 
  749: domU kernels
  750: ------------
  751: 
  752: On a physical computer, the BIOS reads sector 0, and a chain of boot
  753: loaders finds and loads a kernel.  Normally this comes from the root
  754: file system.  With Xen domUs, the process is totally different.  The
  755: normal path is for the domU kernel to be a file in the dom0's
  756: file system.  At the request of the dom0, Xen loads that kernel into a
  757: new domU instance and starts execution.  While domU kernels can be
  758: anyplace, reasonable places to store domU kernels on the dom0 are in /
  759: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
  760: config files), or in /u0/xen (where the vdisks are).
  761: 
  762: Note that loading the domU kernel from the dom0 implies that boot
  763: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
  764: See the VPS section near the end for discussion of alternate ways to
  765: obtain domU kernels.
  766: 
  767: CPU and memory
  768: --------------
  769: 
  770: A domain is provided with some number of vcpus, less than the number
  771: of CPUs seen by the hypervisor.  (For a dom0, this is controlled by
  772: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
  773: from the config file by the "vcpus = N" directive.
  774: 
  775: A domain is provided with memory; this is controlled in the config
  776: file by "memory = N" (in megabytes).  In the straightforward case, the
  777: sum of the the memory allocated to the dom0 and all domUs must be less
  778: than the available memory.
  779: 
  780: Xen also provides a "balloon" driver, which can be used to let domains
  781: use more memory temporarily.  TODO: Explain better, and explain how
  782: well it works with NetBSD.
  783: 
  784: Virtual disks
  785: -------------
  786: 
  787: With the file/vnd style, typically one creates a directory,
  788: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
  789: domUs.  Then, for each domU disk, one writes zeros to a file that then
  790: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
  791: for the first virtual disk for the domU called foo.  Writing zeros to
  792: the file serves two purposes.  One is that preallocating the contents
  793: improves performance.  The other is that vnd on sparse files has
  794: failed to work.  TODO: give working/notworking NetBSD versions for
  795: sparse vnd and gnats reference.  Note that the use of file/vnd for Xen
  796: is not really different than creating a file-backed virtual disk for
  797: some other purpose, except that xentools handles the vnconfig
  798: commands.  To create an empty 4G virtual disk, simply do
  799: 
  800:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
  801: 
  802: Do not use qemu-img-xen, because this will create sparse file.  There
  803: have been recent (2015) reports of sparse vnd(4) devices causing
  804: lockups, but there is apparently no PR.
  805: 
  806: With the lvm style, one creates logical devices.  They are then used
  807: similarly to vnds.  TODO: Add an example with lvm.
  808: 
  809: In domU config files, the disks are defined as a sequence of 3-tuples.
  810: The first element is "method:/path/to/disk".  Common methods are
  811: "file:" for file-backed vnd. and "phy:" for something that is already
  812: a (TODO: character or block) device.
  813: 
  814: The second element is an artifact of how virtual disks are passed to
  815: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
  816: are given a device name to associate with the disk, and values like
  817: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
  818: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
  819: second argument.  The name given is converted to a major/minor by
  820: calling stat(2) on the name in /dev and this is passed to the domU.
  821: In the general case, the dom0 and domU can be different operating
  822: systems, and it is an unwarranted assumption that they have consistent
  823: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
  824: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
  825: for the second works fine and avoids this issue.  For a GNU/Linux
  826: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
  827: /dev/hda1.
  828: 
  829: The third element is "w" for writable disks, and "r" for read-only
  830: disks.
  831: 
  832: Note that NetBSD by default creates only vnd[0123].  If you need more
  833: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
  834: dom0.
  835: 
  836: Note that NetBSD by default creates only xbd[0123].  If you need more
  837: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
  838: 
  839: Virtual Networking
  840: ------------------
  841: 
  842: Xen provides virtual Ethernets, each of which connects the dom0 and a
  843: domU.  For each virtual network, there is an interface "xvifN.M" in
  844: the dom0, and in domU index N, a matching interface xennetM (NetBSD
  845: name).  The interfaces behave as if there is an Ethernet with two
  846: adapters connected.  From this primitive, one can construct various
  847: configurations.  We focus on two common and useful cases for which
  848: there are existing scripts: bridging and NAT.
  849: 
  850: With bridging (in the example above), the domU perceives itself to be
  851: on the same network as the dom0.  For server virtualization, this is
  852: usually best.  Bridging is accomplished by creating a bridge(4) device
  853: and adding the dom0's physical interface and the various xvifN.0
  854: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
  855: config file.  The bridge must be set up already in the dom0; an
  856: example /etc/ifconfig.bridge0 is:
  857: 
  858:         create
  859:         up
  860:         !brconfig bridge0 add wm0
  861: 
  862: With NAT, the domU perceives itself to be behind a NAT running on the
  863: dom0.  This is often appropriate when running Xen on a workstation.
  864: TODO: NAT appears to be configured by "vif = [ '' ]".
  865: 
  866: The MAC address specified is the one used for the interface in the new
  867: domain.  The interface in dom0 will use this address XOR'd with
  868: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
  869: 
  870: Sizing domains
  871: --------------
  872: 
  873: Modern x86 hardware has vast amounts of resources.  However, many
  874: virtual servers can function just fine on far less.  A system with
  875: 256M of RAM and a 4G disk can be a reasonable choice.  Note that it is
  876: far easier to adjust virtual resources than physical ones.  For
  877: memory, it's just a config file edit and a reboot.  For disk, one can
  878: create a new file and vnconfig it (or lvm), and then dump/restore,
  879: just like updating physical disks, but without having to be there and
  880: without those pesky connectors.
  881: 
  882: Starting domains automatically
  883: ------------------------------
  884: 
  885: To start domains foo at bar at boot and shut them down cleanly on dom0
  886: shutdown, in rc.conf add:
  887: 
  888:         xendomains="foo bar"
  889: 
  890: Note that earlier versions of the xentools41 xendomains rc.d script
  891: used xl, when one should use xm with 4.1.
  892: 
  893: Creating specific unprivileged domains (domU)
  894: =============================================
  895: 
  896: Creating domUs is almost entirely independent of operating system.  We
  897: have already presented the basics of config files.  Note that you must
  898: have already completed the dom0 setup so that "xl list" (or "xm list")
  899: works.
  900: 
  901: Creating an unprivileged NetBSD domain (domU)
  902: ---------------------------------------------
  903: 
  904: See the earlier config file, and adjust memory.  Decide on how much
  905: storage you will provide, and prepare it (file or lvm).
  906: 
  907: While the kernel will be obtained from the dom0 file system, the same
  908: file should be present in the domU as /netbsd so that tools like
  909: savecore(8) can work.   (This is helpful but not necessary.)
  910: 
  911: The kernel must be specifically for Xen and for use as a domU.  The
  912: i386 and amd64 provide the following kernels:
  913: 
  914:         i386 XEN3_DOMU
  915:         i386 XEN3PAE_DOMU
  916:         amd64 XEN3_DOMU
  917: 
  918: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
  919: use the PAE version of the i386 kernel.
  920: 
  921: This will boot NetBSD, but this is not that useful if the disk is
  922: empty.  One approach is to unpack sets onto the disk outside of xen
  923: (by mounting it, just as you would prepare a physical disk for a
  924: system you can't run the installer on).
  925: 
  926: A second approach is to run an INSTALL kernel, which has a miniroot
  927: and can load sets from the network.  To do this, copy the INSTALL
  928: kernel to / and change the kernel line in the config file to:
  929: 
  930:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  931: 
  932: Then, start the domain as "xl create -c configname".
  933: 
  934: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
  935: line should be used in the config file.
  936: 
  937:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  938: 
  939: After booting the domain, the option to install via CDROM may be
  940: selected.  The CDROM device should be changed to `xbd1d`.
  941: 
  942: Once done installing, "halt -p" the new domain (don't reboot or halt,
  943: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
  944: config file), switch the config file back to the XEN3_DOMU kernel,
  945: and start the new domain again. Now it should be able to use "root on
  946: xbd0a" and you should have a, functional NetBSD domU.
  947: 
  948: TODO: check if this is still accurate.
  949: When the new domain is booting you'll see some warnings about *wscons*
  950: and the pseudo-terminals. These can be fixed by editing the files
  951: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  952: `/etc/ttys`, except *console*, like this:
  953: 
  954:     console "/usr/libexec/getty Pc"         vt100   on secure
  955:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  956:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  957:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  958:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  959: 
  960: Finally, all screens must be commented out from `/etc/wscons.conf`.
  961: 
  962: It is also desirable to add
  963: 
  964:         powerd=YES
  965: 
  966: in rc.conf. This way, the domain will be properly shut down if
  967: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
  968: 
  969: It is not strictly necessary to have a kernel (as /netbsd) in the domU
  970: file system.  However, various programs (e.g. netstat) will use that
  971: kernel to look up symbols to read from kernel virtual memory.  If
  972: /netbsd is not the running kernel, those lookups will fail.  (This is
  973: not really a Xen-specific issue, but because the domU kernel is
  974: obtained from the dom0, it is far more likely to be out of sync or
  975: missing with Xen.)
  976: 
  977: Creating an unprivileged Linux domain (domU)
  978: --------------------------------------------
  979: 
  980: Creating unprivileged Linux domains isn't much different from
  981: unprivileged NetBSD domains, but there are some details to know.
  982: 
  983: First, the second parameter passed to the disk declaration (the '0x1' in
  984: the example below)
  985: 
  986:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  987: 
  988: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  989: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
  990: So, hda1 which has major 3 and minor 1 on a Linux system will have
  991: device number 0x301.  Alternatively, devices names can be used (hda,
  992: hdb, ...)  as xentools has a table to map these names to devices
  993: numbers.  To export a partition to a Linux guest we can use:
  994: 
  995:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
  996:         root = "/dev/hda1 ro"
  997: 
  998: and it will appear as /dev/hda on the Linux system, and be used as root
  999: partition.
 1000: 
 1001: To install the Linux system on the partition to be exported to the
 1002: guest domain, the following method can be used: install
 1003: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
 1004: that will be the root partition of your Linux domain, and mount it.
 1005: Then copy the files from a working Linux system, make adjustments in
 1006: `/etc` (fstab, network config).  It should also be possible to extract
 1007: binary packages such as .rpm or .deb directly to the mounted partition
 1008: using the appropriate tool, possibly running under NetBSD's Linux
 1009: emulation.  Once the file system has been populated, umount it.  If
 1010: desirable, the file system can be converted to ext3 using tune2fs -j.
 1011: It should now be possible to boot the Linux guest domain, using one of
 1012: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
 1013: 
 1014: To get the Linux console right, you need to add:
 1015: 
 1016:     extra = "xencons=tty1"
 1017: 
 1018: to your configuration since not all Linux distributions auto-attach a
 1019: tty to the xen console.
 1020: 
 1021: Creating an unprivileged Solaris domain (domU)
 1022: ----------------------------------------------
 1023: 
 1024: See possibly outdated
 1025: [Solaris domU instructions](/ports/xen/howto-solaris/).
 1026: 
 1027: 
 1028: PCI passthrough: Using PCI devices in guest domains
 1029: ---------------------------------------------------
 1030: 
 1031: The dom0 can give other domains access to selected PCI
 1032: devices. This can allow, for example, a non-privileged domain to have
 1033: access to a physical network interface or disk controller.  However,
 1034: keep in mind that giving a domain access to a PCI device most likely
 1035: will give the domain read/write access to the whole physical memory,
 1036: as PCs don't have an IOMMU to restrict memory access to DMA-capable
 1037: device.  Also, it's not possible to export ISA devices to non-dom0
 1038: domains, which means that the primary VGA adapter can't be exported.
 1039: A guest domain trying to access the VGA registers will panic.
 1040: 
 1041: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
 1042: not been ported to later versions at this time.
 1043: 
 1044: For a PCI device to be exported to a domU, is has to be attached to
 1045: the "pciback" driver in dom0.  Devices passed to the dom0 via the
 1046: pciback.hide boot parameter will attach to "pciback" instead of the
 1047: usual driver.  The list of devices is specified as "(bus:dev.func)",
 1048: where bus and dev are 2-digit hexadecimal numbers, and func a
 1049: single-digit number:
 1050: 
 1051:         pciback.hide=(00:0a.0)(00:06.0)
 1052: 
 1053: pciback devices should show up in the dom0's boot messages, and the
 1054: devices should be listed in the `/kern/xen/pci` directory.
 1055: 
 1056: PCI devices to be exported to a domU are listed in the "pci" array of
 1057: the domU's config file, with the format "0000:bus:dev.func".
 1058: 
 1059:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
 1060: 
 1061: In the domU an "xpci" device will show up, to which one or more pci
 1062: buses will attach.  Then the PCI drivers will attach to PCI buses as
 1063: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
 1064: or any PCI drivers built in by default; you have to build your own
 1065: kernel to use PCI devices in a domU.  Here's a kernel config example;
 1066: note that only the "xpci" lines are unusual.
 1067: 
 1068:         include         "arch/i386/conf/XEN3_DOMU"
 1069: 
 1070:         # Add support for PCI buses to the XEN3_DOMU kernel
 1071:         xpci* at xenbus ?
 1072:         pci* at xpci ?
 1073: 
 1074:         # PCI USB controllers
 1075:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
 1076: 
 1077:         # USB bus support
 1078:         usb*    at uhci?
 1079: 
 1080:         # USB Hubs
 1081:         uhub*   at usb?
 1082:         uhub*   at uhub? port ? configuration ? interface ?
 1083: 
 1084:         # USB Mass Storage
 1085:         umass*  at uhub? port ? configuration ? interface ?
 1086:         wd*     at umass?
 1087:         # SCSI controllers
 1088:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
 1089: 
 1090:         # SCSI bus support (for both ahc and umass)
 1091:         scsibus* at scsi?
 1092: 
 1093:         # SCSI devices
 1094:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
 1095:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
 1096: 
 1097: 
 1098: NetBSD as a domU in a VPS
 1099: =========================
 1100: 
 1101: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
 1102: hardware.  This section explains how to deal with Xen in a domU as a
 1103: virtual private server where you do not control or have access to the
 1104: dom0.  This is not intended to be an exhaustive list of VPS providers;
 1105: only a few are mentioned that specifically support NetBSD.
 1106: 
 1107: VPS operators provide varying degrees of access and mechanisms for
 1108: configuration.  The big issue is usually how one controls which kernel
 1109: is booted, because the kernel is nominally in the dom0 file system (to
 1110: which VPS users do not normally have access).  A second issue is how
 1111: to install NetBSD.
 1112: A VPS user may want to compile a kernel for security updates, to run
 1113: npf, run IPsec, or any other reason why someone would want to change
 1114: their kernel.
 1115: 
 1116: One approach is to have an administrative interface to upload a kernel,
 1117: or to select from a prepopulated list.  Other approaches are pygrub
 1118: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
 1119: kernel from the domU file system.  This is closer to a regular physical
 1120: computer, where someone who controls a machine can replace the kernel.
 1121: 
 1122: A second issue is multiple CPUs.  With NetBSD 6, domUs support
 1123: multiple vcpus, and it is typical for VPS providers to enable multiple
 1124: CPUs for NetBSD domUs.
 1125: 
 1126: pygrub
 1127: -------
 1128: 
 1129: pygrub runs in the dom0 and looks into the domU file system.  This
 1130: implies that the domU must have a kernel in a file system in a format
 1131: known to pygrub.  As of 2014, pygrub seems to be of mostly historical
 1132: interest.
 1133: 
 1134: pvgrub
 1135: ------
 1136: 
 1137: pvgrub is a version of grub that uses PV operations instead of BIOS
 1138: calls.  It is booted from the dom0 as the domU kernel, and then reads
 1139: /grub/menu.lst and loads a kernel from the domU file system.
 1140: 
 1141: [Panix](http://www.panix.com/) lets users use pvgrub.  Panix reports
 1142: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
 1143: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
 1144: page](http://www.panix.com/v-colo/grub.html), which describes only
 1145: Linux but should be updated to cover NetBSD :-).
 1146: 
 1147: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
 1148: their own kernel.  See then [prgmr.com NetBSD
 1149: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
 1150: (which is in need of updating).
 1151: 
 1152: It appears that [grub's FFS
 1153: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
 1154: does not support all aspects of modern FFS, but there are also reports
 1155: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
 1156: partition for the kernel with the intent that grub can understand it,
 1157: which leads to /netbsd not being the actual kernel.  One must remember
 1158: to update the special boot partition.
 1159: 
 1160: Amazon
 1161: ------
 1162: 
 1163: See the [Amazon EC2 page](../amazon_ec2/).
 1164: 
 1165: Using npf
 1166: ---------
 1167: 
 1168: In standard kernels, npf is a module, and thus cannot be loaded in a
 1169: DOMU kernel.
 1170: 
 1171: TODO: Explain how to compile npf into a custom kernel, answering (but
 1172: note that the problem was caused by not booting the right kernel)
 1173: [this email to
 1174: netbsd-users](https://mail-index.netbsd.org/netbsd-users/2014/12/26/msg015576.html).
 1175: 
 1176: TODO items for improving NetBSD/xen
 1177: ===================================
 1178: 
 1179: * Make the NetBSD dom0 kernel work with SMP.
 1180: * Test the Xen 4.5 packages adequately to be able to recommend them as
 1181:   the standard approach.
 1182: * Get PCI passthrough working on Xen 4.5
 1183: * Get pvgrub into pkgsrc, either via xentools or separately.
 1184: * grub
 1185:   * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
 1186:     fragsize/blocksize (UFS2 support may be present; the point is to
 1187:     make it so that with any UFS1/UFS2 file system setup that works
 1188:     with NetBSD grub will also work).
 1189:     See [pkg/40258](https://gnats.netbsd.org/40258).
 1190:   * Push patches upstream.
 1191:   * Get UFS2 patches into pvgrub.
 1192: * Add support for PV ops to a version of /boot, and make it usable as
 1193:   a kernel in Xen, similar to pvgrub.
 1194: * Solve somehow the issue with modules for GENERIC not being loadable
 1195:   in a Xen dom0 or domU kernel.
 1196: 
 1197: Random pointers
 1198: ===============
 1199: 
 1200: This section contains links from elsewhere not yet integrated into the
 1201: HOWTO, and other guides.
 1202: 
 1203: * http://www.lumbercartel.ca/library/xen/
 1204: * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html
 1205: * https://gmplib.org/~tege/xen.html

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