Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.156

1.144     maxv        1: [[!meta title="Xen HowTo"]]
1.5       mspo        3: Introduction
1.13      gdt         4: ============
1.1       mspo        5: 
1.149     maxv        6: Xen is a hypervisor for x86 hardware, which supports running multiple guest
1.58      gdt         7: operating systems on a single physical machine.  Xen is a Type 1 or
                      8: bare-metal hypervisor; one uses the Xen kernel to control the CPU,
                      9: memory and console, a dom0 operating system which mediates access to
                     10: other hardware (e.g., disks, network, USB), and one or more domU
                     11: operating systems which operate in an unprivileged virtualized
                     12: environment.  IO requests from the domU systems are forwarded by the
1.149     maxv       13: Xen hypervisor to the dom0 to be fulfilled.
1.12      gdt        14: 
1.149     maxv       15: Xen supports different styles of guest:
                     17: [[!table data="""
                     18: Style of guest |Supported by NetBSD
1.150     maxv       19: PV             |Yes (dom0, domU)
                     20: HVM            |Yes (domU)
1.149     maxv       21: PVHVM          |No
                     22: PVH            |No
                     23: """]]
                     25: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
                     26: hardware directly, but instead makes hypercalls to the hypervisor; PV
                     27: guests must be specifically coded for Xen. In HVM mode, no guest
                     28: modification is required; however, hardware support is required, such
                     29: as VT-x on Intel CPUs and SVM on AMD CPUs.
1.27      jnemeth    31: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12      gdt        32: The dom0 can start one or more domUs.  (Booting is explained in detail
                     33: in the dom0 section.)
                     35: This HOWTO presumes a basic familiarity with the Xen system
1.123     gdt        36: architecture, with installing NetBSD on i386/amd64 hardware, and with
                     37: installing software from pkgsrc.  See also the [Xen
                     38: website](
1.1       mspo       39: 
1.150     maxv       40: This HOWTO attempts to address both the case of running a NetBSD dom0
                     41: on hardware and running domUs under it (NetBSD and other), and also
                     42: running NetBSD as a domU in a VPS.
1.15      gdt        44: Versions of Xen and NetBSD
                     45: ==========================
1.27      jnemeth    47: Most of the installation concepts and instructions are independent
                     48: of Xen version and NetBSD version.  This section gives advice on
                     49: which version to choose.  Versions not in pkgsrc and older unsupported
                     50: versions of NetBSD are intentionally ignored.
1.15      gdt        51: 
1.133     gdt        52: The term "amd64" is used to refer to both the NetBSD port and to the
1.150     maxv       53: hardware architecture on which it runs.  Such hardware is generally
                     54: made by both Intel and AMD, and common on PC computers.
1.133     gdt        55: 
1.146     maxv       56: Xen versions
                     57: ------------
1.15      gdt        58: 
1.111     wiz        59: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
1.15      gdt        60: xenkernel and xentools.  We will refer only to the kernel versions,
                     61: but note that both packages must be installed together and must have
                     62: matching versions.
1.145     maxv       64: Versions available in pkgsrc:
1.85      gdt        65: 
1.145     maxv       66: [[!table data="""
                     67: Xen Version    |Package Name   |Xen CPU Support        |EOL'ed By Upstream
                     68: 4.2            |xenkernel42    |32bit, 64bit           |Yes
                     69: 4.5            |xenkernel45    |64bit                  |Yes
                     70: 4.6            |xenkernel46    |64bit                  |Partially
                     71: 4.8            |xenkernel48    |64bit                  |No
                     72: 4.11           |xenkernel411   |64bit                  |No
                     73: """]]
1.113     gdt        74: 
1.96      gdt        75: See also the [Xen Security Advisory page](
1.145     maxv       77: Note: Xen 4.2 was the last version to support 32bit CPUs.
1.147     maxv       79: NetBSD versions
                     80: ---------------
1.15      gdt        81: 
1.147     maxv       82: NetBSD 8 is recommended as the stable version of the most recent
                     83: release for production use.
1.146     maxv       84: 
                     85: For developing Xen, netbsd-current may be appropriate.
1.15      gdt        86: 
                     87: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
                     88: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
1.105     gdt        89: need SMP for dom0 functions; the lack of support is really a problem
                     90: when using a dom0 as a normal computer.)
1.15      gdt        91: 
1.147     maxv       92: Note: NetBSD support is called XEN3. However, it does support Xen 4,
                     93: because the hypercall interface has remained identical.
1.18      gdt        95: Architecture
                     96: ------------
1.155     maxv       98: Xen itself runs on amd64 hardware. Practically, almost any computer
                     99: where one would want to run Xen today supports amd64.
                    101: The dom0 system, plus each domU, can be either i386PAE or amd64.
                    102: i386 without PAE is not supported.
                    104: The standard approach is to use NetBSD/amd64 for the dom0.  For domUs,
                    105: NetBSD/i386 (PAE) and NetBSD/amd64 are in widespread use, and there is
1.135     gdt       106: little to no Xen-specific reason to prefer one over the other.
1.133     gdt       107: 
                    108: Note that to use an i386 dom0 with Xen 4.5 or higher, one must build
1.135     gdt       109: (or obtain from pre-built packages) an amd64 Xen kernel and install
1.154     maxv      110: that on the system.  Almost no one in the NetBSD/Xen community does
                    111: this, and the standard, well-tested, approach is to use an amd64 dom0.
1.135     gdt       112: 
                    113: A [posting on
                    114: xen-devel](
                    115: explained that PV system call overhead was higher on amd64, and thus
                    116: there is some notion that i386 guests are faster.  It goes on to
                    117: caution that the total situation is complex and not entirely
                    118: understood. On top of that caution, the post is about Linux, not
1.155     maxv      119: NetBSD.
1.29      gdt       120: 
1.15      gdt       121: NetBSD as a dom0
                    122: ================
                    124: NetBSD can be used as a dom0 and works very well.  The following
                    125: sections address installation, updating NetBSD, and updating Xen.
1.19      gdt       126: Note that it doesn't make sense to talk about installing a dom0 OS
                    127: without also installing Xen itself.  We first address installing
                    128: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
                    129: NetBSD install to a dom0 install by just changing the kernel and boot
                    130: configuration.
1.15      gdt       131: 
1.45      gdt       132: For experimenting with Xen, a machine with as little as 1G of RAM and
                    133: 100G of disk can work.  For running many domUs in productions, far
1.135     gdt       134: more will be needed; e.g. 4-8G and 1T of disk is reasonable for a
                    135: half-dozen domUs of 512M and 32G each.  Basically, the RAM and disk
                    136: have to be bigger than the sum of the RAM/disk needs of the dom0 and
                    137: all the domUs.
1.45      gdt       138: 
1.142     gdt       139: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
                    140: 512M it worked reliably.  This does not make sense, but if you see
                    141: "not ELF" after Xen boots, try increasing dom0 RAM.
1.15      gdt       143: Styles of dom0 operation
                    144: ------------------------
                    146: There are two basic ways to use Xen.  The traditional method is for
                    147: the dom0 to do absolutely nothing other than providing support to some
                    148: number of domUs.  Such a system was probably installed for the sole
                    149: purpose of hosting domUs, and sits in a server room on a UPS.
                    151: The other way is to put Xen under a normal-usage computer, so that the
                    152: dom0 is what the computer would have been without Xen, perhaps a
                    153: desktop or laptop.  Then, one can run domUs at will.  Purists will
                    154: deride this as less secure than the previous approach, and for a
                    155: computer whose purpose is to run domUs, they are right.  But Xen and a
1.93      gdt       156: dom0 (without domUs) is not meaningfully less secure than the same
1.15      gdt       157: things running without Xen.  One can boot Xen or boot regular NetBSD
                    158: alternately with little problems, simply refraining from starting the
                    159: Xen daemons when not running Xen.
                    161: Note that NetBSD as dom0 does not support multiple CPUs.  This will
1.51      gdt       162: limit the performance of the Xen/dom0 workstation approach.  In theory
                    163: the only issue is that the "backend drivers" are not yet MPSAFE:
1.140     khorben   164:
1.15      gdt       165: 
1.19      gdt       166: Installation of NetBSD
                    167: ----------------------
1.13      gdt       168: 
1.19      gdt       169: First,
1.27      jnemeth   170: [install NetBSD/amd64](/guide/inst/)
1.19      gdt       171: just as you would if you were not using Xen.
                    172: However, the partitioning approach is very important.
                    174: If you want to use RAIDframe for the dom0, there are no special issues
                    175: for Xen.  Typically one provides RAID storage for the dom0, and the
1.22      gdt       176: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
1.111     wiz       177: over a RAID1 header to find /boot from a file system within a RAID
1.22      gdt       178: partition; this is no different when booting Xen.
1.19      gdt       179: 
                    180: There are 4 styles of providing backing storage for the virtual disks
1.93      gdt       181: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
1.19      gdt       182: 
                    183: With raw partitions, one has a disklabel (or gpt) partition sized for
                    184: each virtual disk to be used by the domU.  (If you are able to predict
                    185: how domU usage will evolve, please add an explanation to the HOWTO.
                    186: Seriously, needs tend to change over time.)
1.27      jnemeth   188: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
                    189: for domU disks.  This is almost as efficient as raw disk partitions
                    190: and more flexible.  Hence raw disk partitions should typically not
                    191: be used.
1.19      gdt       192: 
1.111     wiz       193: One can use files in the dom0 file system, typically created by dd'ing
1.19      gdt       194: /dev/zero to create a specific size.  This is somewhat less efficient,
                    195: but very convenient, as one can cp the files for backup, or move them
                    196: between dom0 hosts.
                    198: Finally, in theory one can place the files backing the domU disks in a
                    199: SAN.  (This is an invitation for someone who has done this to add a
                    200: HOWTO page.)
1.1       mspo      201: 
1.19      gdt       202: Installation of Xen
                    203: -------------------
1.1       mspo      204: 
1.155     maxv      205: We will consider that you chose to use Xen 4.8, with NetBSD/amd64 as
                    206: dom0. In the dom0, install xenkernel48 and xentools48 from pkgsrc.
                    207: Ensure that your packages are recent.
                    209: Once this is done, install the Xen kernel itself:
                    211: [[!template id=programlisting text="""
                    212: # cp /usr/pkg/xen48-kernel/xen.gz /
                    213: """]]
                    215: Then, place a NetBSD XEN3_DOM0 kernel in /, copied from
                    216: releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz of a NetBSD build.
1.20      gdt       217: 
1.155     maxv      218: Add a line to /boot.cfg to boot Xen:
1.20      gdt       219: 
1.152     maxv      220: [[!template id=programlisting text="""
                    221: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
                    222: """]]
1.20      gdt       223: 
1.155     maxv      224: which specifies that the dom0 should have 512MB of ram, leaving the rest
                    225: to be allocated for domUs.  To use a serial console, use
1.77      gdt       226: 
1.152     maxv      227: [[!template id=programlisting text="""
                    228: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
                    229: """]]
1.77      gdt       230: 
                    231: which will use the first serial port for Xen (which counts starting
1.139     gson      232: from 1, unlike NetBSD which counts starting from 0), forcing
                    233: speed/parity.  Because the NetBSD command line lacks a
                    234: "console=pc" argument, it will use the default "xencons" console device,
                    235: which directs the console I/O through Xen to the same console device Xen
                    236: itself uses (in this case, the serial port).
1.152     maxv      238: In an attempt to add performance, one can also add:
1.37      gdt       239: 
1.152     maxv      240: [[!template id=programlisting text="""
                    241: dom0_max_vcpus=1 dom0_vcpus_pin
                    242: """]]
1.37      gdt       243: 
                    244: to force only one vcpu to be provided (since NetBSD dom0 can't use
1.155     maxv      245: more) and to pin that vcpu to a physical CPU.
1.20      gdt       246: 
1.93      gdt       247: Xen has [many boot
                    248: options](,
1.111     wiz       249: and other than dom0 memory and max_vcpus, they are generally not
1.93      gdt       250: necessary.
1.20      gdt       252: As with non-Xen systems, you should have a line to boot /netbsd (a
1.155     maxv      253: kernel that works without Xen).
1.127     gdt       254: 
1.76      gdt       255: Now, reboot so that you are running a DOM0 kernel under Xen, rather
                    256: than GENERIC without Xen.
1.54      gdt       258: Using grub (historic)
                    259: ---------------------
                    261: Before NetBSD's native bootloader could support Xen, the use of
                    262: grub was recommended.  If necessary, see the
1.135     gdt       263: [old grub information](/ports/xen/howto-grub).
1.54      gdt       264: 
1.28      gdt       265: The [HowTo on Installing into
1.140     khorben   266: RAID-1](
1.28      gdt       267: explains how to set up booting a dom0 with Xen using grub with
                    268: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
1.135     gdt       269: boot.  Now, just create a system with RAID-1, and alter /boot.cfg as
                    270: described above.)
1.28      gdt       271: 
1.21      gdt       272: Configuring Xen
                    273: ---------------
1.76      gdt       275: Now, you have a system that will boot Xen and the dom0 kernel, but not
                    276: do anything else special.  Make sure that you have rebooted into Xen.
                    277: There will be no domUs, and none can be started because you still have
1.102     gdt       278: to configure the dom0 daemons.
1.21      gdt       279: 
1.102     gdt       280: The daemons which should be run vary with Xen version and with whether
1.152     maxv      281: one is using xm or xl. Xen 4.2 and up packages use xl. To use xm with 4.2,
                    282: edit xendomains to use xm instead.
1.31      gdt       283: 
1.132     gdt       284: For 4.1 and up, you should enable xencommons.  Not enabling xencommons
                    285: will result in a hang; it is necessary to hit ^C on the console to let
                    286: the machine finish booting.  If you are using xm (default in 4.1, or
                    287: if you changed xendomains in 4.2), you should also enable xend:
1.31      gdt       288: 
1.152     maxv      289: [[!template id=programlisting text="""
                    290: xend=YES # only if using xm, and only installed <= 4.2
                    291: xencommons=YES
                    292: """]]
1.31      gdt       293: 
                    294: TODO: Recommend for/against xen-watchdog.
1.27      jnemeth   295: 
1.53      gdt       296: After you have configured the daemons and either started them (in the
1.79      gdt       297: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
1.102     gdt       298: available resources, and running domains.  An example with xl follows:
1.34      gdt       299: 
1.153     maxv      300: [[!template id=programlisting text="""
                    301: # xl dmesg
                    302: ... xen's boot info ...
                    303: # xl info
                    304: ... available memory, etc ...
                    305: # xl list
                    306: Name              Id  Mem(MB)  CPU  State  Time(s)  Console
                    307: Domain-0           0       64    0  r----     58.1
                    308: """]]
                    310: Xen logs will be in /var/log/xen.
1.33      gdt       311: 
1.88      gdt       312: ### Issues with xencommons
                    314: xencommons starts xenstored, which stores data on behalf of dom0 and
                    315: domUs.  It does not currently work to stop and start xenstored.
                    316: Certainly all domUs should be shutdown first, following the sort order
                    317: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
                    318: and is not notified when xenstored exits, leading to not recreating
                    319: the state when the new xenstored starts.  Until there's a mechanism to
                    320: make this work, one should not expect to be able to restart xenstored
                    321: (and thus xencommons).  There is currently no reason to expect that
                    322: this will get fixed any time soon.
1.127     gdt       324: ### No-longer needed advice about devices
                    326: The installation of NetBSD should already have created devices for xen
                    327: (xencons, xenevt, xsd_kva), but if they are not present, create them:
1.153     maxv      329: [[!template id=programlisting text="""
                    330: cd /dev && sh MAKEDEV xen
                    331: """]]
1.127     gdt       332: 
1.41      gdt       333: anita (for testing NetBSD)
                    334: --------------------------
1.82      gdt       336: With the setup so far (assuming 4.2/xl), one should be able to run
                    337: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
                    338: root, because anita must create a domU):
1.153     maxv      340: [[!template id=programlisting text="""
                    341: anita --vmm=xl test file:///usr/obj/i386/
                    342: """]]
1.82      gdt       343: 
                    344: Alternatively, one can use --vmm=xm to use xm-based domU creation
                    345: instead (and must, on Xen <= 4.1).   TODO: confirm that anita xl really works.
1.40      gdt       347: Xen-specific NetBSD issues
                    348: --------------------------
                    350: There are (at least) two additional things different about NetBSD as a
                    351: dom0 kernel compared to hardware.
1.111     wiz       353: One is that the module ABI is different because some of the #defines
1.109     gdt       354: change, so one must build modules for Xen.  As of netbsd-7, the build
                    355: system does this automatically.  TODO: check this.  (Before building
                    356: Xen modules was added, it was awkward to use modules to the point
                    357: where it was considered that it did not work.)
1.40      gdt       358: 
                    359: The other difference is that XEN3_DOM0 does not have exactly the same
                    360: options as GENERIC.  While it is debatable whether or not this is a
                    361: bug, users should be aware of this and can simply add missing config
                    362: items if desired.
1.15      gdt       364: Updating NetBSD in a dom0
                    365: -------------------------
                    367: This is just like updating NetBSD on bare hardware, assuming the new
                    368: version supports the version of Xen you are running.  Generally, one
                    369: replaces the kernel and reboots, and then overlays userland binaries
                    370: and adjusts /etc.
                    372: Note that one must update both the non-Xen kernel typically used for
                    373: rescue purposes and the DOM0 kernel used with Xen.
1.55      gdt       375: Converting from grub to /boot
                    376: -----------------------------
                    378: These instructions were [TODO: will be] used to convert a system from
                    379: grub to /boot.  The system was originally installed in February of
                    380: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
                    381: over time.  Before these commands, it was running NetBSD 6 i386, Xen
                    382: 4.1 and grub, much like the message linked earlier in the grub
                    383: section.
1.111     wiz       385:         # Install MBR bootblocks on both disks. 
1.55      gdt       386:         fdisk -i /dev/rwd0d
                    387:         fdisk -i /dev/rwd1d
                    388:         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
                    389:         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
                    390:         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
                    391:         # Install secondary boot loader
                    392:         cp -p /usr/mdec/boot /
1.111     wiz       393:         # Create boot.cfg following earlier guidance:
1.142     gdt       394:         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
                    395:         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=512M
1.55      gdt       396:         menu=GENERIC:boot
                    397:         menu=GENERIC single-user:boot -s
                    398:         menu=GENERIC.ok:boot netbsd.ok
                    399:         menu=GENERIC.ok single-user:boot netbsd.ok -s
                    400:         menu=Drop to boot prompt:prompt
                    401:         default=1
                    402:         timeout=30
                    404: TODO: actually do this and fix it if necessary.
1.22      gdt       405: 
1.102     gdt       406: Upgrading Xen versions
1.15      gdt       407: ---------------------
1.110     gdt       409: Minor version upgrades are trivial.  Just rebuild/replace the
                    410: xenkernel version and copy the new xen.gz to / (where /boot.cfg
                    411: references it), and reboot.
                    413: Major version upgrades are conceptually not difficult, but can run
                    414: into all the issues found when installing Xen.  Assuming migration
                    415: from 4.1 to 4.2, remove the xenkernel41 and xentools41 packages and
                    416: install the xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz
                    417: to /.
1.21      gdt       418: 
1.102     gdt       419: Ensure that the contents of /etc/rc.d/xen* are correct.  Specifically,
                    420: they must match the package you just installed and not be left over
                    421: from some previous installation.
                    423: Enable the correct set of daemons; see the configuring section above.
                    424: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
                    426: Ensure that the domU config files are valid for the new version.
1.110     gdt       427: Specifically, for 4.x remove autorestart=True, and ensure that disks
                    428: are specified with numbers as the second argument, as the examples
                    429: above show, and not NetBSD device names.
1.15      gdt       430: 
1.28      gdt       431: Unprivileged domains (domU)
                    432: ===========================
                    434: This section describes general concepts about domUs.  It does not
1.33      gdt       435: address specific domU operating systems or how to install them.  The
                    436: config files for domUs are typically in /usr/pkg/etc/xen, and are
1.60      wiki      437: typically named so that the file name, domU name and the domU's host
1.33      gdt       438: name match.
1.111     wiz       440: The domU is provided with CPU and memory by Xen, configured by the
1.33      gdt       441: dom0.  The domU is provided with disk and network by the dom0,
                    442: mediated by Xen, and configured in the dom0.
                    444: Entropy in domUs can be an issue; physical disks and network are on
                    445: the dom0.  NetBSD's /dev/random system works, but is often challenged.
1.48      gdt       447: Config files
                    448: ------------
                    450: There is no good order to present config files and the concepts
                    451: surrounding what is being configured.  We first show an example config
                    452: file, and then in the various sections give details.
                    454: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
                    455: for a large number of well-commented examples, mostly for running
                    456: GNU/Linux.
                    458: The following is an example minimal domain configuration file
                    459: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
                    460: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
                    461: i386 domU).  The domU serves as a network file server.
                    463:         # -*- mode: python; -*-
1.156   ! wiki      465:         name = "domU-id"
1.48      gdt       466:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
                    467:         memory = 1024
                    468:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
                    469:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
                    470:                  'file:/n0/xen/foo-wd1,0x1,w' ]
1.156   ! wiki      472: The domain will have name given in the name setting.  The kernel has the
1.48      gdt       473: host/domU name in it, so that on the dom0 one can update the various
                    474: domUs independently.  The vif line causes an interface to be provided,
                    475: with a specific mac address (do not reuse MAC addresses!), in bridge
                    476: mode.  Two disks are provided, and they are both writable; the bits
                    477: are stored in files and Xen attaches them to a vnd(4) device in the
1.111     wiz       478: dom0 on domain creation.  The system treats xbd0 as the boot device
1.48      gdt       479: without needing explicit configuration.
                    481: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
                    482: that "xm create" takes the name of a config file, while other commands
                    483: take the name of a domain.  To create the domain, connect to the
                    484: console, create the domain while attaching the console, shutdown the
                    485: domain, and see if it has finished stopping, do (or xl with Xen >=
                    486: 4.2):
                    488:         xm create foo
                    489:         xm console foo
                    490:         xm create -c foo
                    491:         xm shutdown foo
1.90      gdt       492:         xm list
1.48      gdt       493: 
                    494: Typing ^] will exit the console session.  Shutting down a domain is
                    495: equivalent to pushing the power button; a NetBSD domU will receive a
                    496: power-press event and do a clean shutdown.  Shutting down the dom0
                    497: will trigger controlled shutdowns of all configured domUs.
                    499: domU kernels
                    500: ------------
                    502: On a physical computer, the BIOS reads sector 0, and a chain of boot
                    503: loaders finds and loads a kernel.  Normally this comes from the root
1.111     wiz       504: file system.  With Xen domUs, the process is totally different.  The
1.48      gdt       505: normal path is for the domU kernel to be a file in the dom0's
1.111     wiz       506: file system.  At the request of the dom0, Xen loads that kernel into a
1.48      gdt       507: new domU instance and starts execution.  While domU kernels can be
                    508: anyplace, reasonable places to store domU kernels on the dom0 are in /
                    509: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
                    510: config files), or in /u0/xen (where the vdisks are).
1.59      gdt       512: Note that loading the domU kernel from the dom0 implies that boot
                    513: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
1.48      gdt       514: See the VPS section near the end for discussion of alternate ways to
                    515: obtain domU kernels.
1.33      gdt       517: CPU and memory
                    518: --------------
1.48      gdt       520: A domain is provided with some number of vcpus, less than the number
1.111     wiz       521: of CPUs seen by the hypervisor.  (For a dom0, this is controlled by
1.48      gdt       522: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
                    523: from the config file by the "vcpus = N" directive.
                    525: A domain is provided with memory; this is controlled in the config
                    526: file by "memory = N" (in megabytes).  In the straightforward case, the
                    527: sum of the the memory allocated to the dom0 and all domUs must be less
1.33      gdt       528: than the available memory.
                    530: Xen also provides a "balloon" driver, which can be used to let domains
                    531: use more memory temporarily.  TODO: Explain better, and explain how
                    532: well it works with NetBSD.
1.28      gdt       533: 
                    534: Virtual disks
                    535: -------------
1.33      gdt       537: With the file/vnd style, typically one creates a directory,
                    538: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
                    539: domUs.  Then, for each domU disk, one writes zeros to a file that then
                    540: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
                    541: for the first virtual disk for the domU called foo.  Writing zeros to
                    542: the file serves two purposes.  One is that preallocating the contents
                    543: improves performance.  The other is that vnd on sparse files has
                    544: failed to work.  TODO: give working/notworking NetBSD versions for
1.127     gdt       545: sparse vnd and gnats reference.  Note that the use of file/vnd for Xen
                    546: is not really different than creating a file-backed virtual disk for
                    547: some other purpose, except that xentools handles the vnconfig
                    548: commands.  To create an empty 4G virtual disk, simply do
1.39      gdt       549: 
                    550:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
1.33      gdt       551: 
1.89      gdt       552: Do not use qemu-img-xen, because this will create sparse file.  There
                    553: have been recent (2015) reports of sparse vnd(4) devices causing
                    554: lockups, but there is apparently no PR.
1.33      gdt       556: With the lvm style, one creates logical devices.  They are then used
1.48      gdt       557: similarly to vnds.  TODO: Add an example with lvm.
                    559: In domU config files, the disks are defined as a sequence of 3-tuples.
                    560: The first element is "method:/path/to/disk".  Common methods are
                    561: "file:" for file-backed vnd. and "phy:" for something that is already
                    562: a (TODO: character or block) device.
                    564: The second element is an artifact of how virtual disks are passed to
                    565: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
                    566: are given a device name to associate with the disk, and values like
                    567: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
                    568: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
                    569: second argument.  The name given is converted to a major/minor by
1.49      gdt       570: calling stat(2) on the name in /dev and this is passed to the domU.
                    571: In the general case, the dom0 and domU can be different operating
1.48      gdt       572: systems, and it is an unwarranted assumption that they have consistent
                    573: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
                    574: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
1.49      gdt       575: for the second works fine and avoids this issue.  For a GNU/Linux
                    576: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
                    577: /dev/hda1.
1.48      gdt       578: 
                    579: The third element is "w" for writable disks, and "r" for read-only
                    580: disks.
1.28      gdt       581: 
1.127     gdt       582: Note that NetBSD by default creates only vnd[0123].  If you need more
                    583: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
                    584: dom0.
                    586: Note that NetBSD by default creates only xbd[0123].  If you need more
                    587: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
1.28      gdt       589: Virtual Networking
                    590: ------------------
1.111     wiz       592: Xen provides virtual Ethernets, each of which connects the dom0 and a
1.46      gdt       593: domU.  For each virtual network, there is an interface "xvifN.M" in
                    594: the dom0, and in domU index N, a matching interface xennetM (NetBSD
                    595: name).  The interfaces behave as if there is an Ethernet with two
1.111     wiz       596: adapters connected.  From this primitive, one can construct various
1.46      gdt       597: configurations.  We focus on two common and useful cases for which
                    598: there are existing scripts: bridging and NAT.
1.28      gdt       599: 
1.48      gdt       600: With bridging (in the example above), the domU perceives itself to be
                    601: on the same network as the dom0.  For server virtualization, this is
                    602: usually best.  Bridging is accomplished by creating a bridge(4) device
                    603: and adding the dom0's physical interface and the various xvifN.0
                    604: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
                    605: config file.  The bridge must be set up already in the dom0; an
                    606: example /etc/ifconfig.bridge0 is:
1.46      gdt       607: 
                    608:         create
                    609:         up
                    610:         !brconfig bridge0 add wm0
1.28      gdt       611: 
                    612: With NAT, the domU perceives itself to be behind a NAT running on the
                    613: dom0.  This is often appropriate when running Xen on a workstation.
1.48      gdt       614: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28      gdt       615: 
1.49      gdt       616: The MAC address specified is the one used for the interface in the new
1.53      gdt       617: domain.  The interface in dom0 will use this address XOR'd with
1.49      gdt       618: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
1.33      gdt       620: Sizing domains
                    621: --------------
                    623: Modern x86 hardware has vast amounts of resources.  However, many
                    624: virtual servers can function just fine on far less.  A system with
1.142     gdt       625: 512M of RAM and a 4G disk can be a reasonable choice.  Note that it is
1.33      gdt       626: far easier to adjust virtual resources than physical ones.  For
                    627: memory, it's just a config file edit and a reboot.  For disk, one can
                    628: create a new file and vnconfig it (or lvm), and then dump/restore,
                    629: just like updating physical disks, but without having to be there and
                    630: without those pesky connectors.
1.48      gdt       632: Starting domains automatically
                    633: ------------------------------
1.28      gdt       634: 
1.48      gdt       635: To start domains foo at bar at boot and shut them down cleanly on dom0
                    636: shutdown, in rc.conf add:
1.28      gdt       637: 
1.48      gdt       638:         xendomains="foo bar"
1.28      gdt       639: 
1.111     wiz       640: Note that earlier versions of the xentools41 xendomains rc.d script
                    641: used xl, when one should use xm with 4.1.
1.28      gdt       642: 
                    643: Creating specific unprivileged domains (domU)
                    644: =============================================
1.14      gdt       645: 
                    646: Creating domUs is almost entirely independent of operating system.  We
1.49      gdt       647: have already presented the basics of config files.  Note that you must
                    648: have already completed the dom0 setup so that "xl list" (or "xm list")
                    649: works.
1.14      gdt       650: 
                    651: Creating an unprivileged NetBSD domain (domU)
                    652: ---------------------------------------------
1.1       mspo      653: 
1.49      gdt       654: See the earlier config file, and adjust memory.  Decide on how much
                    655: storage you will provide, and prepare it (file or lvm).
1.111     wiz       657: While the kernel will be obtained from the dom0 file system, the same
1.49      gdt       658: file should be present in the domU as /netbsd so that tools like
                    659: savecore(8) can work.   (This is helpful but not necessary.)
                    661: The kernel must be specifically for Xen and for use as a domU.  The
                    662: i386 and amd64 provide the following kernels:
                    664:         i386 XEN3PAE_DOMU
1.95      gdt       665:         amd64 XEN3_DOMU
1.5       mspo      666: 
1.49      gdt       667: This will boot NetBSD, but this is not that useful if the disk is
                    668: empty.  One approach is to unpack sets onto the disk outside of xen
                    669: (by mounting it, just as you would prepare a physical disk for a
                    670: system you can't run the installer on).
                    672: A second approach is to run an INSTALL kernel, which has a miniroot
                    673: and can load sets from the network.  To do this, copy the INSTALL
                    674: kernel to / and change the kernel line in the config file to:
1.5       mspo      675: 
1.49      gdt       676:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5       mspo      677: 
1.49      gdt       678: Then, start the domain as "xl create -c configname".
1.1       mspo      679: 
1.49      gdt       680: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
                    681: line should be used in the config file.
1.1       mspo      682: 
1.3       mspo      683:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1       mspo      684: 
                    685: After booting the domain, the option to install via CDROM may be
1.49      gdt       686: selected.  The CDROM device should be changed to `xbd1d`.
1.1       mspo      687: 
1.49      gdt       688: Once done installing, "halt -p" the new domain (don't reboot or halt,
                    689: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
                    690: config file), switch the config file back to the XEN3_DOMU kernel,
                    691: and start the new domain again. Now it should be able to use "root on
                    692: xbd0a" and you should have a, functional NetBSD domU.
1.1       mspo      693: 
1.49      gdt       694: TODO: check if this is still accurate.
1.1       mspo      695: When the new domain is booting you'll see some warnings about *wscons*
                    696: and the pseudo-terminals. These can be fixed by editing the files
1.5       mspo      697: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
                    698: `/etc/ttys`, except *console*, like this:
1.1       mspo      699: 
1.3       mspo      700:     console "/usr/libexec/getty Pc"         vt100   on secure
                    701:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
                    702:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
                    703:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
                    704:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
1.1       mspo      705: 
1.5       mspo      706: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1       mspo      707: 
                    708: It is also desirable to add
1.49      gdt       710:         powerd=YES
1.1       mspo      711: 
1.5       mspo      712: in rc.conf. This way, the domain will be properly shut down if
1.53      gdt       713: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
1.1       mspo      714: 
1.92      gdt       715: It is not strictly necessary to have a kernel (as /netbsd) in the domU
1.111     wiz       716: file system.  However, various programs (e.g. netstat) will use that
1.92      gdt       717: kernel to look up symbols to read from kernel virtual memory.  If
                    718: /netbsd is not the running kernel, those lookups will fail.  (This is
                    719: not really a Xen-specific issue, but because the domU kernel is
                    720: obtained from the dom0, it is far more likely to be out of sync or
                    721: missing with Xen.)
1.14      gdt       723: Creating an unprivileged Linux domain (domU)
1.5       mspo      724: --------------------------------------------
1.1       mspo      725: 
                    726: Creating unprivileged Linux domains isn't much different from
                    727: unprivileged NetBSD domains, but there are some details to know.
                    729: First, the second parameter passed to the disk declaration (the '0x1' in
                    730: the example below)
1.3       mspo      732:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1       mspo      733: 
                    734: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49      gdt       735: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
                    736: So, hda1 which has major 3 and minor 1 on a Linux system will have
                    737: device number 0x301.  Alternatively, devices names can be used (hda,
                    738: hdb, ...)  as xentools has a table to map these names to devices
                    739: numbers.  To export a partition to a Linux guest we can use:
1.1       mspo      740: 
1.49      gdt       741:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
                    742:         root = "/dev/hda1 ro"
1.1       mspo      743: 
                    744: and it will appear as /dev/hda on the Linux system, and be used as root
                    745: partition.
1.49      gdt       747: To install the Linux system on the partition to be exported to the
                    748: guest domain, the following method can be used: install
                    749: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
                    750: that will be the root partition of your Linux domain, and mount it.
                    751: Then copy the files from a working Linux system, make adjustments in
                    752: `/etc` (fstab, network config).  It should also be possible to extract
                    753: binary packages such as .rpm or .deb directly to the mounted partition
                    754: using the appropriate tool, possibly running under NetBSD's Linux
1.111     wiz       755: emulation.  Once the file system has been populated, umount it.  If
                    756: desirable, the file system can be converted to ext3 using tune2fs -j.
1.49      gdt       757: It should now be possible to boot the Linux guest domain, using one of
                    758: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1       mspo      759: 
1.111     wiz       760: To get the Linux console right, you need to add:
1.1       mspo      761: 
1.3       mspo      762:     extra = "xencons=tty1"
1.1       mspo      763: 
1.111     wiz       764: to your configuration since not all Linux distributions auto-attach a
1.1       mspo      765: tty to the xen console.
1.14      gdt       767: Creating an unprivileged Solaris domain (domU)
1.5       mspo      768: ----------------------------------------------
1.1       mspo      769: 
1.50      gdt       770: See possibly outdated
                    771: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5       mspo      772: 
1.1       mspo      773: 
1.52      gdt       774: PCI passthrough: Using PCI devices in guest domains
                    775: ---------------------------------------------------
1.1       mspo      776: 
1.53      gdt       777: The dom0 can give other domains access to selected PCI
1.52      gdt       778: devices. This can allow, for example, a non-privileged domain to have
                    779: access to a physical network interface or disk controller.  However,
                    780: keep in mind that giving a domain access to a PCI device most likely
                    781: will give the domain read/write access to the whole physical memory,
                    782: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53      gdt       783: device.  Also, it's not possible to export ISA devices to non-dom0
1.52      gdt       784: domains, which means that the primary VGA adapter can't be exported.
                    785: A guest domain trying to access the VGA registers will panic.
1.53      gdt       787: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52      gdt       788: not been ported to later versions at this time.
                    790: For a PCI device to be exported to a domU, is has to be attached to
                    791: the "pciback" driver in dom0.  Devices passed to the dom0 via the
                    792: pciback.hide boot parameter will attach to "pciback" instead of the
                    793: usual driver.  The list of devices is specified as "(bus:dev.func)",
1.5       mspo      794: where bus and dev are 2-digit hexadecimal numbers, and func a
                    795: single-digit number:
1.1       mspo      796: 
1.52      gdt       797:         pciback.hide=(00:0a.0)(00:06.0)
1.1       mspo      798: 
1.52      gdt       799: pciback devices should show up in the dom0's boot messages, and the
1.5       mspo      800: devices should be listed in the `/kern/xen/pci` directory.
1.1       mspo      801: 
1.52      gdt       802: PCI devices to be exported to a domU are listed in the "pci" array of
                    803: the domU's config file, with the format "0000:bus:dev.func".
1.1       mspo      804: 
1.52      gdt       805:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1       mspo      806: 
1.52      gdt       807: In the domU an "xpci" device will show up, to which one or more pci
1.111     wiz       808: buses will attach.  Then the PCI drivers will attach to PCI buses as
1.52      gdt       809: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
                    810: or any PCI drivers built in by default; you have to build your own
                    811: kernel to use PCI devices in a domU.  Here's a kernel config example;
                    812: note that only the "xpci" lines are unusual.
                    814:         include         "arch/i386/conf/XEN3_DOMU"
1.111     wiz       816:         # Add support for PCI buses to the XEN3_DOMU kernel
1.52      gdt       817:         xpci* at xenbus ?
                    818:         pci* at xpci ?
                    820:         # PCI USB controllers
                    821:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
                    823:         # USB bus support
                    824:         usb*    at uhci?
                    826:         # USB Hubs
                    827:         uhub*   at usb?
                    828:         uhub*   at uhub? port ? configuration ? interface ?
                    830:         # USB Mass Storage
                    831:         umass*  at uhub? port ? configuration ? interface ?
                    832:         wd*     at umass?
                    833:         # SCSI controllers
                    834:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
                    836:         # SCSI bus support (for both ahc and umass)
                    837:         scsibus* at scsi?
                    839:         # SCSI devices
                    840:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
                    841:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
1.1       mspo      842: 
1.28      gdt       844: NetBSD as a domU in a VPS
                    845: =========================
                    847: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
                    848: hardware.  This section explains how to deal with Xen in a domU as a
                    849: virtual private server where you do not control or have access to the
1.70      gdt       850: dom0.  This is not intended to be an exhaustive list of VPS providers;
                    851: only a few are mentioned that specifically support NetBSD.
1.28      gdt       852: 
1.52      gdt       853: VPS operators provide varying degrees of access and mechanisms for
                    854: configuration.  The big issue is usually how one controls which kernel
1.111     wiz       855: is booted, because the kernel is nominally in the dom0 file system (to
                    856: which VPS users do not normally have access).  A second issue is how
1.70      gdt       857: to install NetBSD.
1.52      gdt       858: A VPS user may want to compile a kernel for security updates, to run
                    859: npf, run IPsec, or any other reason why someone would want to change
                    860: their kernel.
1.111     wiz       862: One approach is to have an administrative interface to upload a kernel,
1.68      gdt       863: or to select from a prepopulated list.  Other approaches are pygrub
1.59      gdt       864: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
1.111     wiz       865: kernel from the domU file system.  This is closer to a regular physical
1.59      gdt       866: computer, where someone who controls a machine can replace the kernel.
1.52      gdt       867: 
1.74      gdt       868: A second issue is multiple CPUs.  With NetBSD 6, domUs support
                    869: multiple vcpus, and it is typical for VPS providers to enable multiple
                    870: CPUs for NetBSD domUs.
1.68      gdt       872: pygrub
1.59      gdt       873: -------
1.52      gdt       874: 
1.111     wiz       875: pygrub runs in the dom0 and looks into the domU file system.  This
                    876: implies that the domU must have a kernel in a file system in a format
1.68      gdt       877: known to pygrub.  As of 2014, pygrub seems to be of mostly historical
                    878: interest.
1.52      gdt       879: 
1.59      gdt       880: pvgrub
                    881: ------
                    883: pvgrub is a version of grub that uses PV operations instead of BIOS
                    884: calls.  It is booted from the dom0 as the domU kernel, and then reads
1.111     wiz       885: /grub/menu.lst and loads a kernel from the domU file system.
1.59      gdt       886: 
1.70      gdt       887: [Panix]( lets users use pvgrub.  Panix reports
1.71      gdt       888: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
                    889: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
1.70      gdt       890: page](, which describes only
1.74      gdt       891: Linux but should be updated to cover NetBSD :-).
1.70      gdt       892: 
                    893: []( also lets users with pvgrub to boot
                    894: their own kernel.  See then [ NetBSD
1.74      gdt       895: HOWTO](
                    896: (which is in need of updating).
1.59      gdt       897: 
1.70      gdt       898: It appears that [grub's FFS
                    899: code](
                    900: does not support all aspects of modern FFS, but there are also reports
1.72      gdt       901: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
1.70      gdt       902: partition for the kernel with the intent that grub can understand it,
                    903: which leads to /netbsd not being the actual kernel.  One must remember
1.111     wiz       904: to update the special boot partition.
1.59      gdt       905: 
                    906: Amazon
                    907: ------
1.143     wiki      909: See the [Amazon EC2 page](/amazon_ec2/).
1.44      gdt       910: 
1.65      gdt       911: TODO items for improving NetBSD/xen
                    912: ===================================
1.93      gdt       914: * Make the NetBSD dom0 kernel work with SMP.
                    915: * Test the Xen 4.5 packages adequately to be able to recommend them as
                    916:   the standard approach.
                    917: * Get PCI passthrough working on Xen 4.5
1.65      gdt       918: * Get pvgrub into pkgsrc, either via xentools or separately.
                    919: * grub
1.70      gdt       920:   * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
1.66      gdt       921:     fragsize/blocksize (UFS2 support may be present; the point is to
1.111     wiz       922:     make it so that with any UFS1/UFS2 file system setup that works
1.66      gdt       923:     with NetBSD grub will also work).
1.140     khorben   924:     See [pkg/40258](
1.65      gdt       925:   * Push patches upstream.
                    926:   * Get UFS2 patches into pvgrub.
                    927: * Add support for PV ops to a version of /boot, and make it usable as
                    928:   a kernel in Xen, similar to pvgrub.
1.93      gdt       929: 
                    930: Random pointers
                    931: ===============
1.117     gdt       933: This section contains links from elsewhere not yet integrated into the
                    934: HOWTO, and other guides.
1.93      gdt       935: 
                    936: *
                    937: *
1.117     gdt       938: *

CVSweb for NetBSD wikisrc <> software: FreeBSD-CVSweb