Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.67

1.5       mspo        1: Introduction
1.13      gdt         2: ============
1.1       mspo        3: 
                      4: [![[Xen
1.57      gdt         5: screenshot]](](
1.1       mspo        6: 
1.58      gdt         7: Xen is a hypervisor (or virtual machine monitor) for x86 hardware
1.12      gdt         8: (i686-class or higher), which supports running multiple guest
1.58      gdt         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.
1.12      gdt        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.
                     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.
1.29      gdt        30: Generally any amd64 machine will work with Xen and PV guests.  In
                     31: theory i386 computers without amd64 support can be used for Xen <=
                     32: 4.2, but we have no recent reports of this working (this is a hint).
                     33: For HVM guests, the VT or VMX cpu feature (Intel) or SVM/HVM/VT
                     34: (amd64) is needed; "cpuctl identify 0" will show this.  TODO: Clean up
                     35: and check the above features.
1.19      gdt        36: 
1.27      jnemeth    37: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12      gdt        38: The dom0 can start one or more domUs.  (Booting is explained in detail
                     39: in the dom0 section.)
                     41: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
                     42: and that Xen kernels and tools are available in pkgsrc.  This HOWTO
                     43: attempts to address both the case of running a NetBSD dom0 on hardware
1.24      gdt        44: and running domUs under it (NetBSD and other), and also running NetBSD
                     45: as a domU in a VPS.
1.12      gdt        46: 
1.20      gdt        47: Some versions of Xen support "PCI passthrough", which means that
                     48: specific PCI devices can be made available to a specific domU instead
                     49: of the dom0.  This can be useful to let a domU run X11, or access some
                     50: network interface or other peripheral.
1.54      gdt        52: NetBSD used to support Xen2; this has been removed.
1.12      gdt        54: Prerequisites
1.13      gdt        55: -------------
1.12      gdt        56: 
                     57: Installing NetBSD/Xen is not extremely difficult, but it is more
                     58: complex than a normal installation of NetBSD.
1.15      gdt        59: In general, this HOWTO is occasionally overly restrictive about how
                     60: things must be done, guiding the reader to stay on the established
                     61: path when there are no known good reasons to stray.
1.12      gdt        62: 
                     63: This HOWTO presumes a basic familiarity with the Xen system
1.16      gdt        64: architecture.  This HOWTO presumes familiarity with installing NetBSD
                     65: on i386/amd64 hardware and installing software from pkgsrc.
1.27      jnemeth    66: See also the [Xen website](
1.1       mspo       67: 
1.15      gdt        68: Versions of Xen and NetBSD
                     69: ==========================
1.27      jnemeth    71: Most of the installation concepts and instructions are independent
                     72: of Xen version and NetBSD version.  This section gives advice on
                     73: which version to choose.  Versions not in pkgsrc and older unsupported
                     74: versions of NetBSD are intentionally ignored.
1.15      gdt        75: 
                     76: Xen
                     77: ---
                     79: In NetBSD, xen is provided in pkgsrc, via matching pairs of packages
                     80: xenkernel and xentools.  We will refer only to the kernel versions,
                     81: but note that both packages must be installed together and must have
                     82: matching versions.
                     84: xenkernel3 and xenkernel33 provide Xen 3.1 and 3.3.  These no longer
1.20      gdt        85: receive security patches and should not be used.  Xen 3.1 supports PCI
1.29      gdt        86: passthrough.  Xen 3.1 supports non-PAE on i386.
1.15      gdt        87: 
                     88: xenkernel41 provides Xen 4.1.  This is no longer maintained by Xen,
                     89: but as of 2014-12 receives backported security patches.  It is a
                     90: reasonable although trailing-edge choice.
                     92: xenkernel42 provides Xen 4.2.  This is maintained by Xen, but old as
                     93: of 2014-12.
                     95: Ideally newer versions of Xen will be added to pkgsrc.
1.26      gdt        97: Note that NetBSD support is called XEN3.  It works with 3.1 through
                     98: 4.2 because the hypercall interface has been stable.
1.20      gdt        99: 
1.19      gdt       100: Xen command program
                    101: -------------------
                    103: Early Xen used a program called "xm" to manipulate the system from the
                    104: dom0.  Starting in 4.1, a replacement program with similar behavior
1.27      jnemeth   105: called "xl" is provided.  In 4.2 and later, "xl" is preferred.  4.4 is
                    106: the last version that has "xm".
1.19      gdt       107: 
1.15      gdt       108: NetBSD
                    109: ------
                    111: The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
                    112: reasonable choices, with more or less the same considerations for
                    113: non-Xen use.  Therefore, netbsd-6 is recommended as the stable version
1.29      gdt       114: of the most recent release for production use.  For those wanting to
                    115: learn Xen or without production stability concerns, netbsd-7 is likely
                    116: most appropriate.
1.15      gdt       117: 
                    118: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
                    119: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
                    120: need SMP; the lack of support is really a problem when using a dom0 as
                    121: a normal computer.)
1.18      gdt       123: Architecture
                    124: ------------
1.29      gdt       126: Xen itself can run on i386 or amd64 machines.  (Practically, almost
                    127: any computer where one would want to run Xen supports amd64.)  If
                    128: using an i386 NetBSD kernel for the dom0, PAE is required (PAE
                    129: versions are built by default).  While i386 dom0 works fine, amd64 is
                    130: recommended as more normal.
                    132: Xen 4.2 is the last version to support i386 as a host.  TODO: Clarify
                    133: if this is about the CPU having to be amd64, or about the dom0 kernel
                    134: having to be amd64.
                    136: One can then run i386 domUs and amd64 domUs, in any combination.  If
                    137: running an i386 NetBSD kernel as a domU, the PAE version is required.
                    138: (Note that emacs (at least) fails if run on i386 with PAE when built
                    139: without, and vice versa, presumably due to bugs in the undump code.)
1.18      gdt       140: 
1.15      gdt       141: Recommendation
                    142: --------------
1.18      gdt       144: Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
1.30      gdt       145: xl, the NetBSD 6 stable branch, and to use an amd64 kernel as the
                    146: dom0.  Either the i386 or amd64 of NetBSD may be used as domUs.
1.15      gdt       147: 
1.36      gdt       148: Build problems
                    149: --------------
                    151: Ideally, all versions of Xen in pkgsrc would build on all versions of
                    152: NetBSD on both i386 and amd64.  However, that isn't the case.  Besides
                    153: aging code and aging compilers, qemu (included in xentools for HVM
1.62      gdt       154: support) is difficult to build.  The following are known to work or FAIL:
1.36      gdt       155: 
1.62      gdt       156:         xenkernel3 netbsd-5 amd64
1.63      gdt       157:         xentools3 netbsd-5 amd64
1.64      gdt       158:         xentools3=hvm netbsd-5 amd64 ????
1.62      gdt       159:         xenkernel33 netbsd-5 amd64
1.63      gdt       160:         xentools33 netbsd-5 amd64
1.36      gdt       161:         xenkernel41 netbsd-5 amd64
                    162:         xentools41 netbsd-5 amd64
1.62      gdt       163:         xenkernel42 netbsd-5 amd64
1.64      gdt       164:         xentools42 netbsd-5 amd64
1.62      gdt       165: 
                    166:         xenkernel3 netbsd-6 i386 FAIL
                    167:         xentools3 netbsd-6 i386
1.63      gdt       168:         xentools3-hvm netbsd-6 i386 FAIL (dependencies fail)
                    169:         xenkernel33 netbsd-6 i386
                    170:         xentools33 netbsd-6 i386
1.36      gdt       171:         xenkernel41 netbsd-6 i386
                    172:         xentools41 netbsd-6 i386
1.63      gdt       173:         xenkernel42 netbsd-6 i386
1.64      gdt       174:         xentools42 netbsd-6 i386 *MIXED
                    176: (*On netbsd-6 i386, there is a xentools42 in the 2014Q3 official builds,
                    177: but it does not build for gdt.)
1.36      gdt       178: 
1.15      gdt       179: NetBSD as a dom0
                    180: ================
                    182: NetBSD can be used as a dom0 and works very well.  The following
                    183: sections address installation, updating NetBSD, and updating Xen.
1.19      gdt       184: Note that it doesn't make sense to talk about installing a dom0 OS
                    185: without also installing Xen itself.  We first address installing
                    186: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
                    187: NetBSD install to a dom0 install by just changing the kernel and boot
                    188: configuration.
1.15      gdt       189: 
1.45      gdt       190: For experimenting with Xen, a machine with as little as 1G of RAM and
                    191: 100G of disk can work.  For running many domUs in productions, far
                    192: more will be needed.
1.15      gdt       194: Styles of dom0 operation
                    195: ------------------------
                    197: There are two basic ways to use Xen.  The traditional method is for
                    198: the dom0 to do absolutely nothing other than providing support to some
                    199: number of domUs.  Such a system was probably installed for the sole
                    200: purpose of hosting domUs, and sits in a server room on a UPS.
                    202: The other way is to put Xen under a normal-usage computer, so that the
                    203: dom0 is what the computer would have been without Xen, perhaps a
                    204: desktop or laptop.  Then, one can run domUs at will.  Purists will
                    205: deride this as less secure than the previous approach, and for a
                    206: computer whose purpose is to run domUs, they are right.  But Xen and a
                    207: dom0 (without domUs) is not meaingfully less secure than the same
                    208: things running without Xen.  One can boot Xen or boot regular NetBSD
                    209: alternately with little problems, simply refraining from starting the
                    210: Xen daemons when not running Xen.
                    212: Note that NetBSD as dom0 does not support multiple CPUs.  This will
1.51      gdt       213: limit the performance of the Xen/dom0 workstation approach.  In theory
                    214: the only issue is that the "backend drivers" are not yet MPSAFE:
1.15      gdt       216: 
1.19      gdt       217: Installation of NetBSD
                    218: ----------------------
1.13      gdt       219: 
1.19      gdt       220: First,
1.27      jnemeth   221: [install NetBSD/amd64](/guide/inst/)
1.19      gdt       222: just as you would if you were not using Xen.
                    223: However, the partitioning approach is very important.
                    225: If you want to use RAIDframe for the dom0, there are no special issues
                    226: for Xen.  Typically one provides RAID storage for the dom0, and the
1.22      gdt       227: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
                    228: over a RAID1 header to find /boot from a filesystem within a RAID
                    229: partition; this is no different when booting Xen.
1.19      gdt       230: 
                    231: There are 4 styles of providing backing storage for the virtual disks
                    232: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN,
                    234: With raw partitions, one has a disklabel (or gpt) partition sized for
                    235: each virtual disk to be used by the domU.  (If you are able to predict
                    236: how domU usage will evolve, please add an explanation to the HOWTO.
                    237: Seriously, needs tend to change over time.)
1.27      jnemeth   239: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
                    240: for domU disks.  This is almost as efficient as raw disk partitions
                    241: and more flexible.  Hence raw disk partitions should typically not
                    242: be used.
1.19      gdt       243: 
                    244: One can use files in the dom0 filesystem, typically created by dd'ing
                    245: /dev/zero to create a specific size.  This is somewhat less efficient,
                    246: but very convenient, as one can cp the files for backup, or move them
                    247: between dom0 hosts.
                    249: Finally, in theory one can place the files backing the domU disks in a
                    250: SAN.  (This is an invitation for someone who has done this to add a
                    251: HOWTO page.)
1.1       mspo      252: 
1.19      gdt       253: Installation of Xen
                    254: -------------------
1.1       mspo      255: 
1.20      gdt       256: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
                    257: pkgsrc (or another matching pair).
                    258: See [the pkgsrc
                    259: documentation]( for help with pkgsrc.
                    261: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
                    262: recent versions have HVM support integrated in the main xentools
                    263: package.  It is entirely reasonable to run only PV guests.
                    265: Next you need to install the selected Xen kernel itself, which is
                    266: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
                    267: For debugging, one may copy xen-debug.gz; this is conceptually similar
                    268: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
                    269: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
                    270: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
                    271: of a NetBSD build.  Both xen and NetBSD may be left compressed.  (If
                    272: using i386, use releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.)
1.49      gdt       274: With Xen as the kernel, you must provide a dom0 NetBSD kernel to be
                    275: used as a module; place this in /.  Suitable kernels are provided in
                    276: releasedir/binary/kernel:
                    278:         i386 XEN3_DOM0
                    279:         i386 XEN3PAE_DOM0
                    280:        amd64 XEN3_DOM0
                    282: The first one is only for use with Xen 3.1 and i386-mode Xen (and you
                    283: should not do this).  Current Xen always uses PAE on i386, but you
                    284: should generally use amd64 for the dom0.  In a dom0 kernel, kernfs is
                    285: mandatory for xend to comunicate with the kernel, so ensure that /kern
                    286: is in fstab.  TODO: Say this is default, or file a PR and give a
                    287: reference.
1.20      gdt       288: 
                    289: Because you already installed NetBSD, you have a working boot setup
                    290: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
                    291: beginning of your root filesystem, /boot present, and likely
                    292: /boot.cfg.  (If not, fix before continuing!)
                    294: See boot.cfg(5) for an example.  The basic line is
1.37      gdt       296:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
1.20      gdt       297: 
                    298: which specifies that the dom0 should have 256M, leaving the rest to be
1.37      gdt       299: allocated for domUs.  In an attempt to add performance, one can also
                    300: add
                    302:         dom0_max_vcpus=1 dom0_vcpus_pin
                    304: to force only one vcpu to be provided (since NetBSD dom0 can't use
                    305: more) and to pin that vcpu to a physical cpu.  TODO: benchmark this.
1.20      gdt       306: 
                    307: As with non-Xen systems, you should have a line to boot /netbsd (a
                    308: kernel that works without Xen) and fallback versions of the non-Xen
                    309: kernel, Xen, and the dom0 kernel.
1.1       mspo      310: 
1.54      gdt       311: Using grub (historic)
                    312: ---------------------
                    314: Before NetBSD's native bootloader could support Xen, the use of
                    315: grub was recommended.  If necessary, see the
                    316: [old grub information](/ports/xen/howto-grub/).
1.28      gdt       318: The [HowTo on Installing into
                    319: RAID-1](
                    320: explains how to set up booting a dom0 with Xen using grub with
                    321: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
                    322: boot.)
1.21      gdt       324: Configuring Xen
                    325: ---------------
1.53      gdt       327: Xen logs will be in /var/log/xen.
1.21      gdt       329: Now, you have a system that will boot Xen and the dom0 kernel, and
                    330: just run the dom0 kernel.  There will be no domUs, and none can be
1.31      gdt       331: started because you still have to configure the dom0 tools.  The
                    332: daemons which should be run vary with Xen version and with whether one
                    333: is using xm or xl.  Note that xend is for supporting "xm", and should
                    334: only be used if you plan on using "xm".  Do NOT enable xend if you
                    335: plan on using "xl" as it will cause problems.
1.21      gdt       336: 
1.43      gdt       337: The installation of NetBSD should already have created devices for xen
                    338: (xencons, xenevt), but if they are not present, create them:
                    340:         cd /dev && sh MAKEDEV xen
1.31      gdt       342: TODO: Give 3.1 advice (or remove it from pkgsrc).
                    344: For 3.3 (and thus xm), add to rc.conf (but note that you should have
                    345: installed 4.1 or 4.2):
1.32      gdt       347:         xend=YES
                    348:         xenbackendd=YES
1.31      gdt       349: 
1.33      gdt       350: For 4.1 (and thus xm; xl is believed not to work well), add to rc.conf:
1.31      gdt       351: 
1.53      gdt       352:         xencommons=YES
1.31      gdt       353:         xend=YES
1.56      gdt       355: (If you are using xentools41 from before 2014-12-26, change
                    356: rc.d/xendomains to use xm rather than xl.)
1.31      gdt       357: 
1.33      gdt       358: For 4.2 with xm, add to rc.conf
1.53      gdt       360:         xencommons=YES
1.33      gdt       361:         xend=YES
                    363: For 4.2 with xl (preferred), add to rc.conf:
1.31      gdt       364: 
1.53      gdt       365:         xencommons=YES
1.31      gdt       366:         TODO: explain if there is a xend replacement
                    368: TODO: Recommend for/against xen-watchdog.
1.27      jnemeth   369: 
1.53      gdt       370: After you have configured the daemons and either started them (in the
                    371: order given) or rebooted, run the following (or use xl) to inspect
                    372: Xen's boot messages, available resources, and running domains:
1.34      gdt       373: 
1.43      gdt       374:         # xm dmesg
                    375:        [xen's boot info]
                    376:         # xm info
                    377:        [available memory, etc.]
                    378:         # xm list
                    379:         Name              Id  Mem(MB)  CPU  State  Time(s)  Console
                    380:         Domain-0           0       64    0  r----     58.1
1.33      gdt       381: 
1.41      gdt       382: anita (for testing NetBSD)
                    383: --------------------------
                    385: With the setup so far, one should be able to run anita (see
                    386: pkgsrc/sysutils/py-anita) to test NetBSD releases, by doing (as root,
                    387: because anita must create a domU):
                    389:         anita --vmm=xm test file:///usr/obj/i386/
                    391: Alternatively, one can use --vmm=xl to use xl-based domU creation instead.
                    392: TODO: check this.
1.40      gdt       394: Xen-specific NetBSD issues
                    395: --------------------------
                    397: There are (at least) two additional things different about NetBSD as a
                    398: dom0 kernel compared to hardware.
                    400: One is that modules are not usable in DOM0 kernels, so one must
                    401: compile in what's needed.  It's not really that modules cannot work,
                    402: but that modules must be built for XEN3_DOM0 because some of the
                    403: defines change and the normal module builds don't do this.  Basically,
                    404: enabling Xen changes the kernel ABI, and the module build system
                    405: doesn't cope with this.
                    407: The other difference is that XEN3_DOM0 does not have exactly the same
                    408: options as GENERIC.  While it is debatable whether or not this is a
                    409: bug, users should be aware of this and can simply add missing config
                    410: items if desired.
1.15      gdt       412: Updating NetBSD in a dom0
                    413: -------------------------
                    415: This is just like updating NetBSD on bare hardware, assuming the new
                    416: version supports the version of Xen you are running.  Generally, one
                    417: replaces the kernel and reboots, and then overlays userland binaries
                    418: and adjusts /etc.
                    420: Note that one must update both the non-Xen kernel typically used for
                    421: rescue purposes and the DOM0 kernel used with Xen.
1.55      gdt       423: Converting from grub to /boot
                    424: -----------------------------
                    426: These instructions were [TODO: will be] used to convert a system from
                    427: grub to /boot.  The system was originally installed in February of
                    428: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
                    429: over time.  Before these commands, it was running NetBSD 6 i386, Xen
                    430: 4.1 and grub, much like the message linked earlier in the grub
                    431: section.
                    433:         # Install mbr bootblocks on both disks. 
                    434:         fdisk -i /dev/rwd0d
                    435:         fdisk -i /dev/rwd1d
                    436:         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
                    437:         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
                    438:         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
                    439:         # Install secondary boot loader
                    440:         cp -p /usr/mdec/boot /
                    441:         # Create boog.cfg following earlier guidance:
                    442:         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
                    443:         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
                    444:         menu=GENERIC:boot
                    445:         menu=GENERIC single-user:boot -s
                    446:         menu=GENERIC.ok:boot netbsd.ok
                    447:         menu=GENERIC.ok single-user:boot netbsd.ok -s
                    448:         menu=Drop to boot prompt:prompt
                    449:         default=1
                    450:         timeout=30
                    452: TODO: actually do this and fix it if necessary.
1.22      gdt       453: 
1.15      gdt       454: Updating Xen versions
                    455: ---------------------
1.21      gdt       457: Updating Xen is conceptually not difficult, but can run into all the
                    458: issues found when installing Xen.  Assuming migration from 4.1 to 4.2,
                    459: remove the xenkernel41 and xentools41 packages and install the
                    460: xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz to /.
                    462: Ensure that the contents of /etc/rc.d/xen* are correct.  Enable the
                    463: correct set of daemons.  Ensure that the domU config files are valid
                    464: for the new version.
1.15      gdt       465: 
1.28      gdt       466: 
                    467: Unprivileged domains (domU)
                    468: ===========================
                    470: This section describes general concepts about domUs.  It does not
1.33      gdt       471: address specific domU operating systems or how to install them.  The
                    472: config files for domUs are typically in /usr/pkg/etc/xen, and are
1.60      wiki      473: typically named so that the file name, domU name and the domU's host
1.33      gdt       474: name match.
                    476: The domU is provided with cpu and memory by Xen, configured by the
                    477: dom0.  The domU is provided with disk and network by the dom0,
                    478: mediated by Xen, and configured in the dom0.
                    480: Entropy in domUs can be an issue; physical disks and network are on
                    481: the dom0.  NetBSD's /dev/random system works, but is often challenged.
1.48      gdt       483: Config files
                    484: ------------
                    486: There is no good order to present config files and the concepts
                    487: surrounding what is being configured.  We first show an example config
                    488: file, and then in the various sections give details.
                    490: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
                    491: for a large number of well-commented examples, mostly for running
                    492: GNU/Linux.
                    494: The following is an example minimal domain configuration file
                    495: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
                    496: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
                    497: i386 domU).  The domU serves as a network file server.
                    499:         # -*- mode: python; -*-
                    501:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
                    502:         memory = 1024
                    503:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
                    504:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
                    505:                  'file:/n0/xen/foo-wd1,0x1,w' ]
                    507: The domain will have the same name as the file.  The kernel has the
                    508: host/domU name in it, so that on the dom0 one can update the various
                    509: domUs independently.  The vif line causes an interface to be provided,
                    510: with a specific mac address (do not reuse MAC addresses!), in bridge
                    511: mode.  Two disks are provided, and they are both writable; the bits
                    512: are stored in files and Xen attaches them to a vnd(4) device in the
                    513: dom0 on domain creation.  The system treates xbd0 as the boot device
                    514: without needing explicit configuration.
                    516: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
                    517: that "xm create" takes the name of a config file, while other commands
                    518: take the name of a domain.  To create the domain, connect to the
                    519: console, create the domain while attaching the console, shutdown the
                    520: domain, and see if it has finished stopping, do (or xl with Xen >=
                    521: 4.2):
                    523:         xm create foo
                    524:         xm console foo
                    525:         xm create -c foo
                    526:         xm shutdown foo
                    527:        xm list
                    529: Typing ^] will exit the console session.  Shutting down a domain is
                    530: equivalent to pushing the power button; a NetBSD domU will receive a
                    531: power-press event and do a clean shutdown.  Shutting down the dom0
                    532: will trigger controlled shutdowns of all configured domUs.
                    534: domU kernels
                    535: ------------
                    537: On a physical computer, the BIOS reads sector 0, and a chain of boot
                    538: loaders finds and loads a kernel.  Normally this comes from the root
                    539: filesystem.  With Xen domUs, the process is totally different.  The
                    540: normal path is for the domU kernel to be a file in the dom0's
                    541: filesystem.  At the request of the dom0, Xen loads that kernel into a
                    542: new domU instance and starts execution.  While domU kernels can be
                    543: anyplace, reasonable places to store domU kernels on the dom0 are in /
                    544: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
                    545: config files), or in /u0/xen (where the vdisks are).
1.59      gdt       547: Note that loading the domU kernel from the dom0 implies that boot
                    548: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
1.48      gdt       549: See the VPS section near the end for discussion of alternate ways to
                    550: obtain domU kernels.
1.33      gdt       552: CPU and memory
                    553: --------------
1.48      gdt       555: A domain is provided with some number of vcpus, less than the number
                    556: of cpus seen by the hypervisor.  (For a dom0, this is controlled by
                    557: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
                    558: from the config file by the "vcpus = N" directive.
                    560: A domain is provided with memory; this is controlled in the config
                    561: file by "memory = N" (in megabytes).  In the straightforward case, the
                    562: sum of the the memory allocated to the dom0 and all domUs must be less
1.33      gdt       563: than the available memory.
                    565: Xen also provides a "balloon" driver, which can be used to let domains
                    566: use more memory temporarily.  TODO: Explain better, and explain how
                    567: well it works with NetBSD.
1.28      gdt       568: 
                    569: Virtual disks
                    570: -------------
1.33      gdt       572: With the file/vnd style, typically one creates a directory,
                    573: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
                    574: domUs.  Then, for each domU disk, one writes zeros to a file that then
                    575: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
                    576: for the first virtual disk for the domU called foo.  Writing zeros to
                    577: the file serves two purposes.  One is that preallocating the contents
                    578: improves performance.  The other is that vnd on sparse files has
                    579: failed to work.  TODO: give working/notworking NetBSD versions for
                    580: sparse vnd.  Note that the use of file/vnd for Xen is not really
                    581: different than creating a file-backed virtual disk for some other
1.39      gdt       582: purpose, except that xentools handles the vnconfig commands.  To
                    583: create an empty 4G virtual disk, simply do
                    585:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
1.33      gdt       586: 
                    587: With the lvm style, one creates logical devices.  They are then used
1.48      gdt       588: similarly to vnds.  TODO: Add an example with lvm.
                    590: In domU config files, the disks are defined as a sequence of 3-tuples.
                    591: The first element is "method:/path/to/disk".  Common methods are
                    592: "file:" for file-backed vnd. and "phy:" for something that is already
                    593: a (TODO: character or block) device.
                    595: The second element is an artifact of how virtual disks are passed to
                    596: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
                    597: are given a device name to associate with the disk, and values like
                    598: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
                    599: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
                    600: second argument.  The name given is converted to a major/minor by
1.49      gdt       601: calling stat(2) on the name in /dev and this is passed to the domU.
                    602: In the general case, the dom0 and domU can be different operating
1.48      gdt       603: systems, and it is an unwarranted assumption that they have consistent
                    604: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
                    605: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
1.49      gdt       606: for the second works fine and avoids this issue.  For a GNU/Linux
                    607: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
                    608: /dev/hda1.
1.48      gdt       609: 
                    610: The third element is "w" for writable disks, and "r" for read-only
                    611: disks.
1.28      gdt       612: 
                    613: Virtual Networking
                    614: ------------------
1.46      gdt       616: Xen provides virtual ethernets, each of which connects the dom0 and a
                    617: domU.  For each virtual network, there is an interface "xvifN.M" in
                    618: the dom0, and in domU index N, a matching interface xennetM (NetBSD
                    619: name).  The interfaces behave as if there is an Ethernet with two
                    620: adaptors connected.  From this primitive, one can construct various
                    621: configurations.  We focus on two common and useful cases for which
                    622: there are existing scripts: bridging and NAT.
1.28      gdt       623: 
1.48      gdt       624: With bridging (in the example above), the domU perceives itself to be
                    625: on the same network as the dom0.  For server virtualization, this is
                    626: usually best.  Bridging is accomplished by creating a bridge(4) device
                    627: and adding the dom0's physical interface and the various xvifN.0
                    628: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
                    629: config file.  The bridge must be set up already in the dom0; an
                    630: example /etc/ifconfig.bridge0 is:
1.46      gdt       631: 
                    632:         create
                    633:         up
                    634:         !brconfig bridge0 add wm0
1.28      gdt       635: 
                    636: With NAT, the domU perceives itself to be behind a NAT running on the
                    637: dom0.  This is often appropriate when running Xen on a workstation.
1.48      gdt       638: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28      gdt       639: 
1.49      gdt       640: The MAC address specified is the one used for the interface in the new
1.53      gdt       641: domain.  The interface in dom0 will use this address XOR'd with
1.49      gdt       642: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
1.33      gdt       644: Sizing domains
                    645: --------------
                    647: Modern x86 hardware has vast amounts of resources.  However, many
                    648: virtual servers can function just fine on far less.  A system with
                    649: 256M of RAM and a 4G disk can be a reasonable choice.  Note that it is
                    650: far easier to adjust virtual resources than physical ones.  For
                    651: memory, it's just a config file edit and a reboot.  For disk, one can
                    652: create a new file and vnconfig it (or lvm), and then dump/restore,
                    653: just like updating physical disks, but without having to be there and
                    654: without those pesky connectors.
1.48      gdt       656: Starting domains automatically
                    657: ------------------------------
1.28      gdt       658: 
1.48      gdt       659: To start domains foo at bar at boot and shut them down cleanly on dom0
                    660: shutdown, in rc.conf add:
1.28      gdt       661: 
1.48      gdt       662:         xendomains="foo bar"
1.28      gdt       663: 
                    664: TODO: Explain why 4.1 rc.d/xendomains has xl, when one should use xm
1.48      gdt       665: on 4.1.  Or fix the xentools41 package to have xm
1.28      gdt       666: 
                    667: Creating specific unprivileged domains (domU)
                    668: =============================================
1.14      gdt       669: 
                    670: Creating domUs is almost entirely independent of operating system.  We
1.49      gdt       671: have already presented the basics of config files.  Note that you must
                    672: have already completed the dom0 setup so that "xl list" (or "xm list")
                    673: works.
1.14      gdt       674: 
                    675: Creating an unprivileged NetBSD domain (domU)
                    676: ---------------------------------------------
1.1       mspo      677: 
1.49      gdt       678: See the earlier config file, and adjust memory.  Decide on how much
                    679: storage you will provide, and prepare it (file or lvm).
                    681: While the kernel will be obtained from the dom0 filesystem, the same
                    682: file should be present in the domU as /netbsd so that tools like
                    683: savecore(8) can work.   (This is helpful but not necessary.)
                    685: The kernel must be specifically for Xen and for use as a domU.  The
                    686: i386 and amd64 provide the following kernels:
                    688:         i386 XEN3_DOMU
                    689:         i386 XEN3PAE_DOMU
                    690:        amd64 XEN3_DOMU
1.5       mspo      691: 
1.49      gdt       692: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
                    693: use the PAE version of the i386 kernel.
                    695: This will boot NetBSD, but this is not that useful if the disk is
                    696: empty.  One approach is to unpack sets onto the disk outside of xen
                    697: (by mounting it, just as you would prepare a physical disk for a
                    698: system you can't run the installer on).
                    700: A second approach is to run an INSTALL kernel, which has a miniroot
                    701: and can load sets from the network.  To do this, copy the INSTALL
                    702: kernel to / and change the kernel line in the config file to:
1.5       mspo      703: 
1.49      gdt       704:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5       mspo      705: 
1.49      gdt       706: Then, start the domain as "xl create -c configname".
1.1       mspo      707: 
1.49      gdt       708: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
                    709: line should be used in the config file.
1.1       mspo      710: 
1.3       mspo      711:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1       mspo      712: 
                    713: After booting the domain, the option to install via CDROM may be
1.49      gdt       714: selected.  The CDROM device should be changed to `xbd1d`.
1.1       mspo      715: 
1.49      gdt       716: Once done installing, "halt -p" the new domain (don't reboot or halt,
                    717: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
                    718: config file), switch the config file back to the XEN3_DOMU kernel,
                    719: and start the new domain again. Now it should be able to use "root on
                    720: xbd0a" and you should have a, functional NetBSD domU.
1.1       mspo      721: 
1.49      gdt       722: TODO: check if this is still accurate.
1.1       mspo      723: When the new domain is booting you'll see some warnings about *wscons*
                    724: and the pseudo-terminals. These can be fixed by editing the files
1.5       mspo      725: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
                    726: `/etc/ttys`, except *console*, like this:
1.1       mspo      727: 
1.3       mspo      728:     console "/usr/libexec/getty Pc"         vt100   on secure
                    729:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
                    730:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
                    731:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
                    732:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
1.1       mspo      733: 
1.5       mspo      734: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1       mspo      735: 
                    736: It is also desirable to add
1.49      gdt       738:         powerd=YES
1.1       mspo      739: 
1.5       mspo      740: in rc.conf. This way, the domain will be properly shut down if
1.53      gdt       741: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
1.1       mspo      742: 
                    743: Your domain should be now ready to work, enjoy.
1.14      gdt       745: Creating an unprivileged Linux domain (domU)
1.5       mspo      746: --------------------------------------------
1.1       mspo      747: 
                    748: Creating unprivileged Linux domains isn't much different from
                    749: unprivileged NetBSD domains, but there are some details to know.
                    751: First, the second parameter passed to the disk declaration (the '0x1' in
                    752: the example below)
1.3       mspo      754:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1       mspo      755: 
                    756: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49      gdt       757: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
                    758: So, hda1 which has major 3 and minor 1 on a Linux system will have
                    759: device number 0x301.  Alternatively, devices names can be used (hda,
                    760: hdb, ...)  as xentools has a table to map these names to devices
                    761: numbers.  To export a partition to a Linux guest we can use:
1.1       mspo      762: 
1.49      gdt       763:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
                    764:         root = "/dev/hda1 ro"
1.1       mspo      765: 
                    766: and it will appear as /dev/hda on the Linux system, and be used as root
                    767: partition.
1.49      gdt       769: To install the Linux system on the partition to be exported to the
                    770: guest domain, the following method can be used: install
                    771: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
                    772: that will be the root partition of your Linux domain, and mount it.
                    773: Then copy the files from a working Linux system, make adjustments in
                    774: `/etc` (fstab, network config).  It should also be possible to extract
                    775: binary packages such as .rpm or .deb directly to the mounted partition
                    776: using the appropriate tool, possibly running under NetBSD's Linux
                    777: emulation.  Once the filesystem has been populated, umount it.  If
                    778: desirable, the filesystem can be converted to ext3 using tune2fs -j.
                    779: It should now be possible to boot the Linux guest domain, using one of
                    780: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1       mspo      781: 
                    782: To get the linux console right, you need to add:
1.3       mspo      784:     extra = "xencons=tty1"
1.1       mspo      785: 
                    786: to your configuration since not all linux distributions auto-attach a
                    787: tty to the xen console.
1.14      gdt       789: Creating an unprivileged Solaris domain (domU)
1.5       mspo      790: ----------------------------------------------
1.1       mspo      791: 
1.50      gdt       792: See possibly outdated
                    793: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5       mspo      794: 
1.1       mspo      795: 
1.52      gdt       796: PCI passthrough: Using PCI devices in guest domains
                    797: ---------------------------------------------------
1.1       mspo      798: 
1.53      gdt       799: The dom0 can give other domains access to selected PCI
1.52      gdt       800: devices. This can allow, for example, a non-privileged domain to have
                    801: access to a physical network interface or disk controller.  However,
                    802: keep in mind that giving a domain access to a PCI device most likely
                    803: will give the domain read/write access to the whole physical memory,
                    804: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53      gdt       805: device.  Also, it's not possible to export ISA devices to non-dom0
1.52      gdt       806: domains, which means that the primary VGA adapter can't be exported.
                    807: A guest domain trying to access the VGA registers will panic.
1.53      gdt       809: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52      gdt       810: not been ported to later versions at this time.
                    812: For a PCI device to be exported to a domU, is has to be attached to
                    813: the "pciback" driver in dom0.  Devices passed to the dom0 via the
                    814: pciback.hide boot parameter will attach to "pciback" instead of the
                    815: usual driver.  The list of devices is specified as "(bus:dev.func)",
1.5       mspo      816: where bus and dev are 2-digit hexadecimal numbers, and func a
                    817: single-digit number:
1.1       mspo      818: 
1.52      gdt       819:         pciback.hide=(00:0a.0)(00:06.0)
1.1       mspo      820: 
1.52      gdt       821: pciback devices should show up in the dom0's boot messages, and the
1.5       mspo      822: devices should be listed in the `/kern/xen/pci` directory.
1.1       mspo      823: 
1.52      gdt       824: PCI devices to be exported to a domU are listed in the "pci" array of
                    825: the domU's config file, with the format "0000:bus:dev.func".
1.1       mspo      826: 
1.52      gdt       827:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1       mspo      828: 
1.52      gdt       829: In the domU an "xpci" device will show up, to which one or more pci
                    830: busses will attach.  Then the PCI drivers will attach to PCI busses as
                    831: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
                    832: or any PCI drivers built in by default; you have to build your own
                    833: kernel to use PCI devices in a domU.  Here's a kernel config example;
                    834: note that only the "xpci" lines are unusual.
                    836:         include         "arch/i386/conf/XEN3_DOMU"
                    838:         # Add support for PCI busses to the XEN3_DOMU kernel
                    839:         xpci* at xenbus ?
                    840:         pci* at xpci ?
                    842:         # PCI USB controllers
                    843:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
                    845:         # USB bus support
                    846:         usb*    at uhci?
                    848:         # USB Hubs
                    849:         uhub*   at usb?
                    850:         uhub*   at uhub? port ? configuration ? interface ?
                    852:         # USB Mass Storage
                    853:         umass*  at uhub? port ? configuration ? interface ?
                    854:         wd*     at umass?
                    855:         # SCSI controllers
                    856:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
                    858:         # SCSI bus support (for both ahc and umass)
                    859:         scsibus* at scsi?
                    861:         # SCSI devices
                    862:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
                    863:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
1.1       mspo      864: 
1.28      gdt       866: NetBSD as a domU in a VPS
                    867: =========================
                    869: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
                    870: hardware.  This section explains how to deal with Xen in a domU as a
                    871: virtual private server where you do not control or have access to the
                    872: dom0.
1.52      gdt       874: VPS operators provide varying degrees of access and mechanisms for
                    875: configuration.  The big issue is usually how one controls which kernel
                    876: is booted, because the kernel is nominally in the dom0 filesystem (to
                    877: which VPS users do not normally have acesss).
                    879: A VPS user may want to compile a kernel for security updates, to run
                    880: npf, run IPsec, or any other reason why someone would want to change
                    881: their kernel.
                    883: One approach is to have an adminstrative interface to upload a kernel,
1.59      gdt       884: or to select from a prepopulated list.  Other approaches are py-grub
                    885: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
                    886: kernel from the domU filesystem.  This is closer to a regular physical
                    887: computer, where someone who controls a machine can replace the kernel.
1.52      gdt       888: 
1.59      gdt       889: py-grub
                    890: -------
1.52      gdt       891: 
1.59      gdt       892: py-grub runs in the dom0 and looks into the domU filesystem.  This
                    893: implies that the domU must have a kernel in a filesystem in a format
                    894: known to py-grub.  As of 2014, py-grub seems to be of mostly historical interest.
1.52      gdt       895: 
1.59      gdt       896: pvgrub
                    897: ------
                    899: pvgrub is a version of grub that uses PV operations instead of BIOS
                    900: calls.  It is booted from the dom0 as the domU kernel, and then reads
                    901: /grub/menu.lst and loads a kernel from the domU filesystem.
                    903: []( uses this approach to let users choose
                    904: their own operating system and kernel.  See then [ NetBSD
                    905: HOWTO](
1.61      gdt       907: Because [grub's FFS code](
                    908: appears not to support all aspects of modern FFS,
                    909: typically one has an ext2 or FAT partition for the kernel, so that
1.59      gdt       910: grub can understand it, which leads to /netbsd not being the actual
                    911: kernel.  One must remember to update the special boot partiion.
                    913: Amazon
                    914: ------
                    916: TODO: add link to NetBSD amazon howto.
1.44      gdt       917: 
                    918: Using npf
                    919: ---------
                    921: In standard kernels, npf is a module, and thus cannot be loadeed in a
                    922: DOMU kernel.
1.59      gdt       924: TODO: explain how to compile npf into a custom kernel, answering (but
                    925: note that the problem was caused by not booting the right kernel):
1.44      gdt       926:
1.65      gdt       927: 
                    928: TODO items for improving NetBSD/xen
                    929: ===================================
                    931: * Package Xen 4.4.
1.67    ! gdt       932: * Get PCI passthrough working on Xen 4.2 (or 4.4).
1.65      gdt       933: * Get pvgrub into pkgsrc, either via xentools or separately.
                    934: * grub
1.66      gdt       935:   * Add support to pkgsrc grub2 for UFS2 and arbitrary
                    936:     fragsize/blocksize (UFS2 support may be present; the point is to
                    937:     make it so that with any UFS1/UFS2 filesystem setup that works
                    938:     with NetBSD grub will also work).
1.65      gdt       939:   * Push patches upstream.
                    940:   * Get UFS2 patches into pvgrub.
                    941: * Add support for PV ops to a version of /boot, and make it usable as
                    942:   a kernel in Xen, similar to pvgrub.

CVSweb for NetBSD wikisrc <> software: FreeBSD-CVSweb