Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.201

1.201   ! gdt         1: [[!meta title="Xen HowTo and Status"]]
1.144     maxv        2: 
1.157     maxv        3: Xen is a Type 1 hypervisor which supports running multiple guest operating
                      4: systems on a single physical machine. One uses the Xen kernel to control the
                      5: CPU, memory and console, a dom0 operating system which mediates access to
                      6: other hardware (e.g., disks, network, USB), and one or more domU operating
                      7: systems which operate in an unprivileged virtualized environment. IO requests
                      8: from the domU systems are forwarded by the Xen hypervisor to the dom0 to be
                      9: fulfilled.
1.12      gdt        10: 
1.182     gdt        11: This HOWTO presumes a basic familiarity with the Xen system
                     12: architecture, with installing NetBSD on amd64 hardware, and with
                     13: installing software from pkgsrc.  See also the [Xen
                     14: website](
1.201   ! gdt        16: This document also provides status on what Xen things work on NetBSD
        !            17: (upstream documentation might say something works if it works on some
        !            18: particular Linux system).
        !            19: 
1.182     gdt        20: [[!toc]]
1.183     gdt        22: # Overview
1.182     gdt        23: 
                     24: The basic concept of Xen is that the hypervisor (xenkernel) runs on
                     25: the hardware, and runs a privileged domain ("dom0") that can access
                     26: disks/networking/etc.  One then runs additonal unprivileged domains
                     27: (each a "domU"), presumably to do something useful.
                     29: This HOWTO addresses how to run a NetBSD dom0 (and hence also build
                     30: xen itself).  It also addresses how to run domUs in that environment,
                     31: and how to deal with having a domU in a Xen environment run by someone
                     32: else and/or not running NetBSD.
1.186     gdt        34: There are many choices one can make; the HOWTO recommends the standard
                     35: approach and limits discussion of alternatives in many cases.
1.182     gdt        37: ## Guest Styles
                     39: Xen supports different styles of guests.
1.149     maxv       40: 
                     41: [[!table data="""
                     42: Style of guest |Supported by NetBSD
1.150     maxv       43: PV             |Yes (dom0, domU)
                     44: HVM            |Yes (domU)
1.171     gdt        45: PVHVM          |current-only (domU)
1.180     gdt        46: PVH            |current-only (domU, dom0 not yet)
1.149     maxv       47: """]]
                     49: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
                     50: hardware directly, but instead makes hypercalls to the hypervisor; PV
1.170     gdt        51: guests must be specifically coded for Xen.
1.182     gdt        52: See [PV](\)).
1.170     gdt        53: 
                     54: In HVM mode, no guest modification is required; however, hardware
                     55: support is required, such as VT-x on Intel CPUs and SVM on AMD CPUs.
1.199     gdt        56: The dom0 runs qemu to emulate hardware.  It is therefore non-sensical
                     57: to have an HVM dom0.
1.170     gdt        58: 
1.176     gdt        59: In PVHVM mode, the guest runs as HVM, but additionally can use PV
1.199     gdt        60: drivers for efficiency.  Therefore it is non-sensical for to have a
                     61: PVHVM dom0.  See [PV on HVM](
1.170     gdt        62: 
1.180     gdt        63: There have been two PVH modes: original PVH and PVHv2.  Original PVH
                     64: was based on PV mode and is no longer relevant at all.  PVHv2 is
                     65: basically lightweight HVM with PV drivers.  A critical feature of it
                     66: is that qemu is not needed; the hypervisor can do the emulation that
                     67: is required.  Thus, a dom0 can be PVHv2.
1.182     gdt        68: The source code uses PVH and config files use pvh; this refers to PVHv2.
                     69: See [PVH(v2)](\)_Domu).
1.180     gdt        70: 
1.182     gdt        71: At system boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12      gdt        72: The dom0 can start one or more domUs.  (Booting is explained in detail
                     73: in the dom0 section.)
1.182     gdt        75: ## CPU Architecture
                     77: Xen runs on x86_64 hardware (the NetBSD amd64 port).
                     79: There is a concept of Xen running on ARM, but there are no reports of this working with NetBSD.
                     81: The dom0 system should be amd64.  (Instructions for i386PAE dom0 have been removed from the HOWTO.)
1.1       mspo       82: 
1.182     gdt        83: The domU can be i386PAE or amd64.
                     84: i386PAE at one point was considered as [faster]( than amd64.
1.150     maxv       85: 
1.182     gdt        86: ## Xen Versions
1.15      gdt        87: 
1.111     wiz        88: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
1.15      gdt        89: xenkernel and xentools.  We will refer only to the kernel versions,
                     90: but note that both packages must be installed together and must have
                     91: matching versions.
1.145     maxv       93: Versions available in pkgsrc:
1.85      gdt        94: 
1.145     maxv       95: [[!table data="""
1.184     gdt        96: Xen Version    |Package Name   |Xen CPU Support        |EOL'ed By Upstream
                     97: 4.11           |xenkernel411   |x86_64                 |No
                     98: 4.13           |xenkernel413   |x86_64                 |No
1.145     maxv       99: """]]
1.113     gdt       100: 
1.96      gdt       101: See also the [Xen Security Advisory page](
1.166     gdt       103: Older Xen had a python-based management tool called xm, now replaced
1.176     gdt       104: by xl.
1.166     gdt       105: 
1.183     gdt       106: ## NetBSD versions
                    108: Xen has been supported in NetBSD for a long time, at least since 2005.
                    109: Initially Xen was PV only.
                    111: NetBSD 8 and up support PV and HVM modes.
                    113: Support for PVHVM and PVH is available only in NetBSD-current.
1.191     gdt       115: NetBSD up to and including NetBSD 9 as a dom0 does not run SMP,
                    116: because some drivers are not yet safe for this.  NetBSD-current
                    117: supports SMP in dom0.
1.185     gdt       118: 
                    119: NetBSD, when run as a domU, can and does typically run SMP.
1.191     gdt       121: Note that while Xen 4.13 is current, the kernel support is still
                    122: called XEN3, because the hypercall interface has not changed
                    123: significantly.
1.185     gdt       124: 
1.187     gdt       125: # Creating a NetBSD dom0
1.15      gdt       126: 
1.191     gdt       127: In order to install a NetBSD as a dom0, one first installs a normal
                    128: NetBSD system, and then pivot the install to a dom0 install by
                    129: changing the kernel and boot configuration.
1.15      gdt       130: 
1.142     gdt       131: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
                    132: 512M it worked reliably.  This does not make sense, but if you see
                    133: "not ELF" after Xen boots, try increasing dom0 RAM.
1.181     gdt       135: ## Installation of NetBSD
1.13      gdt       136: 
1.191     gdt       137: [Install NetBSD/amd64](/guide/inst/) just as you would if you were not
                    138: using Xen.  Therefore, use the most recent release, or a build from
                    139: the most recent stable branch.  Alternatively, use -current, being
                    140: mindful of all the usual caveats of lower stability of current, and
                    141: likely a bit more so.
1.1       mspo      142: 
1.181     gdt       143: ## Installation of Xen
1.1       mspo      144: 
1.191     gdt       145: ### Building Xen
1.187     gdt       147: Use the most recent version of Xen in pkgsrc, unless the DESCR says that it is not suitable.
1.186     gdt       148: Therefore, choose 4.13.
                    149: In the dom0, install xenkernel413 and xentools413 from pkgsrc.
1.155     maxv      150: 
1.186     gdt       151: Once this is done, copy the Xen kernel from where pkgsrc puts it to
                    152: where the boot process will be able to find it:
1.155     maxv      153: 
                    154: [[!template id=programlisting text="""
1.186     gdt       155: # cp -p /usr/pkg/xen413-kernel/xen.gz /
1.155     maxv      156: """]]
1.191     gdt       158: Then, place a NetBSD XEN3_DOM0 kernel in the `/` directory. Such
                    159: kernel can either be taken from a local release run, compiled
                    160: manually, or downloaded from the NetBSD FTP, for example at:
1.159     maxv      161: 
                    162: [[!template id=programlisting text="""
1.186     gdt       163:
1.159     maxv      164: """]]
1.20      gdt       165: 
1.191     gdt       166: ### Configuring booting
1.198     gdt       168: Read boot.cfg(8) carefully.  Add lines to /boot.cfg to boot Xen,
                    169: adjusting for your root filesystem:
1.20      gdt       170: 
1.157     maxv      171: [[!template id=filecontent name="/boot.cfg" text="""
1.198     gdt       172: menu=Xen:load /netbsd-XEN3_DOM0.gz root=wd0a console=pc;multiboot /xen.gz dom0_mem=512M
                    173: menu=Xen single user:load /netbsd-XEN3_DOM0.gz root=wd0a console=pc -s;multiboot /xen.gz dom0_mem=512M
1.152     maxv      174: """]]
1.20      gdt       175: 
1.159     maxv      176: This specifies that the dom0 should have 512MB of ram, leaving the rest
1.187     gdt       177: to be allocated for domUs.
1.191     gdt       179: NB: This says add, not replace, so that you will be able to more
                    180: easily boot a NetBSD kernel without Xen.  Once Xen boots ok, you may
                    181: want to set it as default.  It is highly likely that you will have
                    182: trouble at some point, and keeping an up-to-date GENERIC for use in
                    183: fixing problems is the standard prudent approach.
1.194     gdt       185: \todo Explain why rndseed is not set with Xen as part of the dom0
                    186: subconfiguration.
1.198     gdt       188: Note that you are likely to have to set root= because the boot device
                    189: from /boot is not passed via Xen to the dom0 kernel.  With one disk,
                    190: it will work, but e.g. plugging in USB disk to a machine with root on
                    191: wd0a causes boot to fail.
1.195     gdt       193: Beware that userconf statements must be attached to the dom0 load, and
                    194: may not be at top-level, because then they would try to configure the
                    195: hypervisor, if there is a way to pass them via multiboot .  It appears
                    196: that adding `userconf=pckbc` to `/boot.cfg` causes Xen to crash very
                    197: early with a heap overflow.
1.191     gdt       199: ### Console selection
                    201: See boot_console(8).  Understand that you should start from a place of
                    202: having console setup correct for booting GENERIC before trying to
                    203: configure Xen.
1.193     gdt       205: Generally for GENERIC, one sets the console in bootxx_ffsv1 or
                    206: equivalent, and this is passed on to /boot (where one typically does
                    207: not set the console).  This configuration of bootxx_ffsv1 should also
                    208: be in place for Xen systems, to allow seeing messages from /boot and
                    209: use of a keyboard to select a line from the menu.  And, one should
                    210: have a working boot path to GENERIC for rescue situations.
                    212: With GENERIC, the boot options are passed on to /netbsd, but there is
                    213: currently no mechanism to pass these via multiboot to the hypervisor.
                    214: Thus, in addition to configuring the console in the boot blocks, one
                    215: must also configure it for Xen.
                    217: By default, the hypervisor (Xen itself) will use some sort of vga
                    218: device as the console, much like GENERIC uses by default.  The vga
                    219: console is relinquished at the conclusion of hypervisor boot, before
1.194     gdt       220: the dom0 is started.  Xen when using a vga console does not process
                    221: console input.
1.77      gdt       222: 
1.193     gdt       223: The hypervisor can be configured to use a serial port console, e.g.
1.157     maxv      224: [[!template id=filecontent name="/boot.cfg" text="""
1.193     gdt       225: menu=Xen:losad /netbsd-XEN3_DOM0.gz console=com0;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
1.152     maxv      226: """]]
1.194     gdt       227: This example uses the first serial port (Xen counts from 1; this is
1.193     gdt       228: what NetBSD would call com0), and sets speed and parity.  (The dom0 is
                    229: then configured to use the same serial port in this example.)
1.194     gdt       231: With the hypervisor configured for a serial console, it can get input,
                    232: and there is a notion of passing this input to the dom0.  \todo
                    233: Explain why, if Xen has a serial console, the dom0 console is
                    234: typically also configured to open that same serial port, instead of
                    235: getting the passthrough input via the xen console.
1.193     gdt       237: One also configures the console for the dom0.  While one might expect
                    238: console=pc to be default, following behavior of GENERIC, a hasty read
                    239: of the code suggests there is no default and booting without a
                    240: selected console might lead to a panic.  Also, there is merit in
                    241: explicit configuration.  Therefore the standard approach is to place
                    242: console=pc as part of the load statement for the dom0 kernel, or
                    243: alternatively console=com0.
                    245: The NetBSD dom0 kernel will attach xencons(4) (the man page does not
                    246: exist), but this is not used as a console.  It is used to obtain the
                    247: messages from the hypervisor's console; run `xl dmesg` to see them.
1.191     gdt       248: 
                    249: ### Tuning
1.157     maxv      251: In an attempt to add performance, one can also add `dom0_max_vcpus=1 dom0_vcpus_pin`,
1.37      gdt       252: to force only one vcpu to be provided (since NetBSD dom0 can't use
1.158     maxv      253: more) and to pin that vcpu to a physical CPU. Xen has
1.178     gdt       254: [many boot options](,
1.111     wiz       255: and other than dom0 memory and max_vcpus, they are generally not
1.93      gdt       256: necessary.
1.191     gdt       257: \todo Revisit this advice with current.
                    258: \todo Explain if anyone has ever actually measured that this helps.
                    260: ### rc.conf
1.93      gdt       261: 
1.186     gdt       262: Ensure that the boot scripts installed in
                    263: `/usr/pkg/share/examples/rc.d` are in `/etc/rc.d`, either because you
                    264: have `PKG_RCD_SCRIPTS=yes`, or manually.  (This is not special to Xen,
                    265: but a normal part of pkgsrc usage.)
1.159     maxv      266: 
1.186     gdt       267: Set `xencommons=YES` in rc.conf:
1.31      gdt       268: 
1.157     maxv      269: [[!template id=filecontent name="/etc/rc.conf" text="""
1.152     maxv      270: xencommons=YES
                    271: """]]
1.31      gdt       272: 
1.187     gdt       273: \todo Recommend for/against xen-watchdog.
1.191     gdt       275: ### Testing
1.158     maxv      277: Now, reboot so that you are running a DOM0 kernel under Xen, rather
                    278: than GENERIC without Xen.
1.157     maxv      279: 
1.158     maxv      280: Once the reboot is done, use `xl` to inspect Xen's boot messages,
1.157     maxv      281: available resources, and running domains.  For example:
1.34      gdt       282: 
1.153     maxv      283: [[!template id=programlisting text="""
                    284: # xl dmesg
                    285: ... xen's boot info ...
                    286: # xl info
                    287: ... available memory, etc ...
                    288: # xl list
                    289: Name              Id  Mem(MB)  CPU  State  Time(s)  Console
                    290: Domain-0           0       64    0  r----     58.1
                    291: """]]
                    293: Xen logs will be in /var/log/xen.
1.33      gdt       294: 
1.88      gdt       295: ### Issues with xencommons
1.157     maxv      297: `xencommons` starts `xenstored`, which stores data on behalf of dom0 and
1.88      gdt       298: domUs.  It does not currently work to stop and start xenstored.
                    299: Certainly all domUs should be shutdown first, following the sort order
                    300: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
                    301: and is not notified when xenstored exits, leading to not recreating
                    302: the state when the new xenstored starts.  Until there's a mechanism to
                    303: make this work, one should not expect to be able to restart xenstored
                    304: (and thus xencommons).  There is currently no reason to expect that
                    305: this will get fixed any time soon.
1.187     gdt       306: \todo Confirm if this is still true in 2020.
1.82      gdt       307: 
1.181     gdt       308: ## Xen-specific NetBSD issues
1.40      gdt       309: 
                    310: There are (at least) two additional things different about NetBSD as a
                    311: dom0 kernel compared to hardware.
1.191     gdt       313: One is that through NetBSD 9 the module ABI is different because some
                    314: of the #defines change, so there are separate sets of modules in
                    315: /stand.  In NetBSD-current, there is only one set of modules.
1.40      gdt       316: 
                    317: The other difference is that XEN3_DOM0 does not have exactly the same
                    318: options as GENERIC.  While it is debatable whether or not this is a
                    319: bug, users should be aware of this and can simply add missing config
                    320: items if desired.
1.187     gdt       322: Finally, there have been occasional reports of trouble with X11
                    323: servers in NetBSD as a dom0.
                    325: ## Updating Xen in a dom0
                    327: Basically, update the xenkernel and xentools packages and copy the new
                    328: Xen kernel into place, and reboot.  This procedure should be usable to
                    329: update to a new Xen release, but the reader is reminded that having a
                    330: non-Xen boot methods was recommended earlier.
1.181     gdt       332: ## Updating NetBSD in a dom0
1.15      gdt       333: 
                    334: This is just like updating NetBSD on bare hardware, assuming the new
                    335: version supports the version of Xen you are running.  Generally, one
                    336: replaces the kernel and reboots, and then overlays userland binaries
1.157     maxv      337: and adjusts `/etc`.
1.15      gdt       338: 
1.191     gdt       339: Note that one should update both the non-Xen kernel typically used for
                    340: rescue purposes, as well as the DOM0 kernel used with Xen.
1.15      gdt       341: 
1.187     gdt       342: ## anita (for testing NetBSD)
                    344: With a NetBSD dom0, even without any domUs, one should be able to run
                    345: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
                    346: root, because anita must create a domU):
                    348: [[!template id=programlisting text="""
                    349: anita --vmm=xl test file:///usr/obj/i386/
                    350: """]]
                    352: # Unprivileged domains (domU)
1.28      gdt       353: 
                    354: This section describes general concepts about domUs.  It does not
1.33      gdt       355: address specific domU operating systems or how to install them.  The
1.157     maxv      356: config files for domUs are typically in `/usr/pkg/etc/xen`, and are
1.60      wiki      357: typically named so that the file name, domU name and the domU's host
1.33      gdt       358: name match.
1.111     wiz       360: The domU is provided with CPU and memory by Xen, configured by the
1.33      gdt       361: dom0.  The domU is provided with disk and network by the dom0,
                    362: mediated by Xen, and configured in the dom0.
                    364: Entropy in domUs can be an issue; physical disks and network are on
                    365: the dom0.  NetBSD's /dev/random system works, but is often challenged.
1.181     gdt       367: ## Config files
1.48      gdt       368: 
1.200     gdt       369: See /usr/pkg/share/examples/xen/xlexample* for a very small number of
                    370: examples for running GNU/Linux.
1.48      gdt       371: 
1.157     maxv      372: The following is an example minimal domain configuration file. The domU
                    373: serves as a network file server.
                    375: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
                    376: name = "domU-id"
                    377: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
                    378: memory = 1024
                    379: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
                    380: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
                    381:          'file:/n0/xen/foo-wd1,0x1,w' ]
                    382: """]]
1.48      gdt       383: 
1.157     maxv      384: The domain will have name given in the `name` setting.  The kernel has the
1.48      gdt       385: host/domU name in it, so that on the dom0 one can update the various
1.157     maxv      386: domUs independently.  The `vif` line causes an interface to be provided,
1.48      gdt       387: with a specific mac address (do not reuse MAC addresses!), in bridge
                    388: mode.  Two disks are provided, and they are both writable; the bits
                    389: are stored in files and Xen attaches them to a vnd(4) device in the
1.111     wiz       390: dom0 on domain creation.  The system treats xbd0 as the boot device
1.48      gdt       391: without needing explicit configuration.
1.200     gdt       393: There is not type line; that implicitly defines a pv domU.  Otherwise,
                    394: one sets type to the lower-case version of the domU type in the table
1.201   ! gdt       395: above, e.g. `type = "hvm"`.
1.187     gdt       396: 
1.163     gson      397: By convention, domain config files are kept in `/usr/pkg/etc/xen`.  Note
1.157     maxv      398: that "xl create" takes the name of a config file, while other commands
                    399: take the name of a domain.
                    401: Examples of commands:
1.48      gdt       402: 
1.157     maxv      403: [[!template id=programlisting text="""
1.163     gson      404: xl create /usr/pkg/etc/xen/foo
                    405: xl console domU-id
                    406: xl create -c /usr/pkg/etc/xen/foo
                    407: xl shutdown domU-id
1.157     maxv      408: xl list
                    409: """]]
                    411: Typing `^]` will exit the console session.  Shutting down a domain is
1.48      gdt       412: equivalent to pushing the power button; a NetBSD domU will receive a
                    413: power-press event and do a clean shutdown.  Shutting down the dom0
                    414: will trigger controlled shutdowns of all configured domUs.
1.181     gdt       416: ## CPU and memory
1.33      gdt       417: 
1.169     plunky    418: A domain is provided with some number of vcpus, up to the number
1.157     maxv      419: of CPUs seen by the hypervisor. For a domU, it is controlled
1.48      gdt       420: from the config file by the "vcpus = N" directive.
                    422: A domain is provided with memory; this is controlled in the config
                    423: file by "memory = N" (in megabytes).  In the straightforward case, the
                    424: sum of the the memory allocated to the dom0 and all domUs must be less
1.33      gdt       425: than the available memory.
1.197     gdt       427: ## Balloon driver
                    429: Xen provides a `balloon` driver, which can be used to let domains use
                    430: more memory temporarily.
                    432: \todo Explain how to set up a aystem to use the balloon scheme in a
                    433: useful manner.
1.28      gdt       434: 
1.181     gdt       435: ## Virtual disks
1.28      gdt       436: 
1.158     maxv      437: In domU config files, the disks are defined as a sequence of 3-tuples:
                    439:  * The first element is "method:/path/to/disk". Common methods are
1.162     gutterid  440:    "file:" for a file-backed vnd, and "phy:" for something that is already
                    441:    a device, such as an LVM logical volume.
1.158     maxv      442: 
                    443:  * The second element is an artifact of how virtual disks are passed to
                    444:    Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
                    445:    are given a device name to associate with the disk, and values like
                    446:    "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
                    447:    as xbd0, the second as xbd1, and so on.  However, xl demands a
                    448:    second argument.  The name given is converted to a major/minor by
                    449:    calling stat(2) on the name in /dev and this is passed to the domU.
                    450:    In the general case, the dom0 and domU can be different operating
                    451:    systems, and it is an unwarranted assumption that they have consistent
                    452:    numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
                    453:    as both dom0 and domU, using values of 0x0 for the first disk and 0x1
                    454:    for the second works fine and avoids this issue.  For a GNU/Linux
                    455:    guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
                    456:    /dev/hda1.
1.48      gdt       457: 
1.158     maxv      458:  * The third element is "w" for writable disks, and "r" for read-only
                    459:    disks.
                    461: Example:
                    462: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
                    463: disk = [ 'file:/n0/xen/foo-wd0,0x0,w' ]
                    464: """]]
1.28      gdt       465: 
1.127     gdt       466: Note that NetBSD by default creates only vnd[0123].  If you need more
                    467: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
                    468: dom0.
1.187     gdt       470: ## Virtual Networking
1.28      gdt       471: 
1.111     wiz       472: Xen provides virtual Ethernets, each of which connects the dom0 and a
1.46      gdt       473: domU.  For each virtual network, there is an interface "xvifN.M" in
1.158     maxv      474: the dom0, and a matching interface xennetM (NetBSD name) in domU index N.
                    475: The interfaces behave as if there is an Ethernet with two
1.111     wiz       476: adapters connected.  From this primitive, one can construct various
1.46      gdt       477: configurations.  We focus on two common and useful cases for which
                    478: there are existing scripts: bridging and NAT.
1.28      gdt       479: 
1.48      gdt       480: With bridging (in the example above), the domU perceives itself to be
                    481: on the same network as the dom0.  For server virtualization, this is
                    482: usually best.  Bridging is accomplished by creating a bridge(4) device
                    483: and adding the dom0's physical interface and the various xvifN.0
                    484: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
                    485: config file.  The bridge must be set up already in the dom0; an
                    486: example /etc/ifconfig.bridge0 is:
1.46      gdt       487: 
1.157     maxv      488: [[!template id=filecontent name="/etc/ifconfig.bridge0" text="""
                    489: create
                    490: up
                    491: !brconfig bridge0 add wm0
                    492: """]]
1.28      gdt       493: 
                    494: With NAT, the domU perceives itself to be behind a NAT running on the
                    495: dom0.  This is often appropriate when running Xen on a workstation.
1.48      gdt       496: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28      gdt       497: 
1.49      gdt       498: The MAC address specified is the one used for the interface in the new
1.53      gdt       499: domain.  The interface in dom0 will use this address XOR'd with
1.49      gdt       500: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
1.187     gdt       502: ## Starting domains automatically
1.28      gdt       503: 
1.158     maxv      504: To start domains `domU-netbsd` and `domU-linux` at boot and shut them
                    505: down cleanly on dom0 shutdown, add the following in rc.conf:
1.28      gdt       506: 
1.158     maxv      507: [[!template id=filecontent name="/etc/rc.conf" text="""
                    508: xendomains="domU-netbsd domU-linux"
                    509: """]]
1.28      gdt       510: 
1.188     gdt       511: # domU setup for specific systems
1.14      gdt       512: 
                    513: Creating domUs is almost entirely independent of operating system.  We
1.188     gdt       514: have already presented the basics of config files in the previous system.
                    516: Of course, this section presumes that you have a working dom0.
1.14      gdt       517: 
1.187     gdt       518: ## Creating a NetBSD PV domU
1.49      gdt       520: See the earlier config file, and adjust memory.  Decide on how much
1.162     gutterid  521: storage you will provide, and prepare it (file or LVM).
1.49      gdt       522: 
1.111     wiz       523: While the kernel will be obtained from the dom0 file system, the same
1.49      gdt       524: file should be present in the domU as /netbsd so that tools like
                    525: savecore(8) can work.   (This is helpful but not necessary.)
1.188     gdt       527: The kernel must be specifically built for Xen, to use PV interfacesas
                    528: a domU.  NetBSD release builds provide the following kernels:
1.49      gdt       529: 
                    530:         i386 XEN3PAE_DOMU
1.95      gdt       531:         amd64 XEN3_DOMU
1.5       mspo      532: 
1.49      gdt       533: This will boot NetBSD, but this is not that useful if the disk is
1.188     gdt       534: empty.  One approach is to unpack sets onto the disk outside of Xen
1.49      gdt       535: (by mounting it, just as you would prepare a physical disk for a
                    536: system you can't run the installer on).
                    538: A second approach is to run an INSTALL kernel, which has a miniroot
                    539: and can load sets from the network.  To do this, copy the INSTALL
                    540: kernel to / and change the kernel line in the config file to:
1.5       mspo      541: 
1.49      gdt       542:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5       mspo      543: 
1.163     gson      544: Then, start the domain as "xl create -c configfile".
1.1       mspo      545: 
1.49      gdt       546: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
                    547: line should be used in the config file.
1.1       mspo      548: 
1.3       mspo      549:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1       mspo      550: 
                    551: After booting the domain, the option to install via CDROM may be
1.49      gdt       552: selected.  The CDROM device should be changed to `xbd1d`.
1.1       mspo      553: 
1.188     gdt       554: Once done installing, "halt -p" the new domain (don't reboot or halt:
1.49      gdt       555: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
                    556: config file), switch the config file back to the XEN3_DOMU kernel,
                    557: and start the new domain again. Now it should be able to use "root on
1.188     gdt       558: xbd0a" and you should have a functional NetBSD domU.
1.1       mspo      559: 
1.49      gdt       560: TODO: check if this is still accurate.
1.1       mspo      561: When the new domain is booting you'll see some warnings about *wscons*
                    562: and the pseudo-terminals. These can be fixed by editing the files
1.5       mspo      563: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
                    564: `/etc/ttys`, except *console*, like this:
1.1       mspo      565: 
1.3       mspo      566:     console "/usr/libexec/getty Pc"         vt100   on secure
                    567:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
                    568:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
                    569:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
                    570:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
1.1       mspo      571: 
1.5       mspo      572: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1       mspo      573: 
1.192     gdt       574: One should also run `powerd` in a domU, but this should not need
                    575: configuring.  With powerd, the domain will run a controlled shutdown
                    576: if `xl shutdown -R` or `xl shutdown -H` is used on the dom0, via
                    577: receiving a synthetic `power button pressed` signal.  In 9 and
                    578: current, `powerd` is run by default under Xen kernels (or if ACPI is
                    579: present), and it can be added to rc.conf if not.
1.1       mspo      580: 
1.92      gdt       581: It is not strictly necessary to have a kernel (as /netbsd) in the domU
1.111     wiz       582: file system.  However, various programs (e.g. netstat) will use that
1.92      gdt       583: kernel to look up symbols to read from kernel virtual memory.  If
                    584: /netbsd is not the running kernel, those lookups will fail.  (This is
                    585: not really a Xen-specific issue, but because the domU kernel is
                    586: obtained from the dom0, it is far more likely to be out of sync or
                    587: missing with Xen.)
1.187     gdt       589: Note that NetBSD by default creates only xbd[0123].  If you need more
                    590: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
1.196     gdt       592: ## Creating a Linux PV domU
1.1       mspo      593: 
                    594: Creating unprivileged Linux domains isn't much different from
                    595: unprivileged NetBSD domains, but there are some details to know.
                    597: First, the second parameter passed to the disk declaration (the '0x1' in
                    598: the example below)
1.3       mspo      600:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1       mspo      601: 
                    602: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49      gdt       603: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
                    604: So, hda1 which has major 3 and minor 1 on a Linux system will have
                    605: device number 0x301.  Alternatively, devices names can be used (hda,
                    606: hdb, ...)  as xentools has a table to map these names to devices
                    607: numbers.  To export a partition to a Linux guest we can use:
1.1       mspo      608: 
1.49      gdt       609:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
                    610:         root = "/dev/hda1 ro"
1.1       mspo      611: 
                    612: and it will appear as /dev/hda on the Linux system, and be used as root
                    613: partition.
1.49      gdt       615: To install the Linux system on the partition to be exported to the
                    616: guest domain, the following method can be used: install
                    617: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
                    618: that will be the root partition of your Linux domain, and mount it.
                    619: Then copy the files from a working Linux system, make adjustments in
                    620: `/etc` (fstab, network config).  It should also be possible to extract
                    621: binary packages such as .rpm or .deb directly to the mounted partition
                    622: using the appropriate tool, possibly running under NetBSD's Linux
1.111     wiz       623: emulation.  Once the file system has been populated, umount it.  If
                    624: desirable, the file system can be converted to ext3 using tune2fs -j.
1.49      gdt       625: It should now be possible to boot the Linux guest domain, using one of
                    626: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1       mspo      627: 
1.111     wiz       628: To get the Linux console right, you need to add:
1.1       mspo      629: 
1.3       mspo      630:     extra = "xencons=tty1"
1.1       mspo      631: 
1.111     wiz       632: to your configuration since not all Linux distributions auto-attach a
1.1       mspo      633: tty to the xen console.
1.180     gdt       635: ## Creating a NetBSD HVM domU
1.196     gdt       637: Use type='hvm', probably.  Use a GENERIC kernel within the disk image.
1.180     gdt       638: 
                    639: ## Creating a NetBSD PVH domU
1.196     gdt       641: This only works with a current kernel in the domU.
1.180     gdt       642: 
1.196     gdt       643: Use type='pvh'.  Probably, use a GENERIC kernel within the disk image,
                    644: which in current has PV support.
                    646: \todo Verify.
                    648: \todo Verify if one can have current PVH domU on a 9 dom0.
1.180     gdt       649: 
1.187     gdt       650: ## Creating a Solaris domU
1.1       mspo      651: 
1.50      gdt       652: See possibly outdated
                    653: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5       mspo      654: 
1.187     gdt       655: ## PCI passthrough: Using PCI devices in guest domains
1.1       mspo      656: 
1.180     gdt       657: NB: PCI passthrough only works on some Xen versions and as of 2020 it
1.196     gdt       658: is not clear that it works on any version in pkgsrc.  \todo Reports
1.180     gdt       659: confirming or denying this notion should be sent to port-xen@.
1.53      gdt       661: The dom0 can give other domains access to selected PCI
1.52      gdt       662: devices. This can allow, for example, a non-privileged domain to have
                    663: access to a physical network interface or disk controller.  However,
                    664: keep in mind that giving a domain access to a PCI device most likely
                    665: will give the domain read/write access to the whole physical memory,
                    666: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53      gdt       667: device.  Also, it's not possible to export ISA devices to non-dom0
1.52      gdt       668: domains, which means that the primary VGA adapter can't be exported.
                    669: A guest domain trying to access the VGA registers will panic.
1.53      gdt       671: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52      gdt       672: not been ported to later versions at this time.
                    674: For a PCI device to be exported to a domU, is has to be attached to
                    675: the "pciback" driver in dom0.  Devices passed to the dom0 via the
                    676: pciback.hide boot parameter will attach to "pciback" instead of the
                    677: usual driver.  The list of devices is specified as "(bus:dev.func)",
1.5       mspo      678: where bus and dev are 2-digit hexadecimal numbers, and func a
                    679: single-digit number:
1.1       mspo      680: 
1.52      gdt       681:         pciback.hide=(00:0a.0)(00:06.0)
1.1       mspo      682: 
1.52      gdt       683: pciback devices should show up in the dom0's boot messages, and the
1.5       mspo      684: devices should be listed in the `/kern/xen/pci` directory.
1.1       mspo      685: 
1.52      gdt       686: PCI devices to be exported to a domU are listed in the "pci" array of
                    687: the domU's config file, with the format "0000:bus:dev.func".
1.1       mspo      688: 
1.52      gdt       689:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1       mspo      690: 
1.52      gdt       691: In the domU an "xpci" device will show up, to which one or more pci
1.111     wiz       692: buses will attach.  Then the PCI drivers will attach to PCI buses as
1.52      gdt       693: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
                    694: or any PCI drivers built in by default; you have to build your own
                    695: kernel to use PCI devices in a domU.  Here's a kernel config example;
                    696: note that only the "xpci" lines are unusual.
                    698:         include         "arch/i386/conf/XEN3_DOMU"
1.111     wiz       700:         # Add support for PCI buses to the XEN3_DOMU kernel
1.52      gdt       701:         xpci* at xenbus ?
                    702:         pci* at xpci ?
                    704:         # PCI USB controllers
                    705:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
                    707:         # USB bus support
                    708:         usb*    at uhci?
                    710:         # USB Hubs
                    711:         uhub*   at usb?
                    712:         uhub*   at uhub? port ? configuration ? interface ?
                    714:         # USB Mass Storage
                    715:         umass*  at uhub? port ? configuration ? interface ?
                    716:         wd*     at umass?
                    717:         # SCSI controllers
                    718:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
                    720:         # SCSI bus support (for both ahc and umass)
                    721:         scsibus* at scsi?
                    723:         # SCSI devices
                    724:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
                    725:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
1.1       mspo      726: 
1.189     gdt       728: # Miscellaneous Information
1.173     gdt       729: 
1.189     gdt       730: ## Nesting under Linux KVM
                    732: It is possible to run a Xen and a NetBSD dom0 under Linux KVM.  One
                    733: can enable virtio in the dom0 for greater speed.
                    735: ## Other nesting
                    737: In theory, any full emulation should be able to run Xen and a NetBSD
                    738: dom0.  The HOWTO does not currently have information about Xen XVM
                    739: mode, nvmm, qemu, Virtualbox, etc.
                    741: ## NetBSD 5 as domU
1.173     gdt       742: 
1.176     gdt       743: [NetBSD 5 is known to panic.](
1.175     gdt       744: (However, NetBSD 5 systems should be updated to a supported version.)
1.173     gdt       745: 
                    746: # NetBSD as a domU in a VPS
1.28      gdt       747: 
                    748: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
                    749: hardware.  This section explains how to deal with Xen in a domU as a
                    750: virtual private server where you do not control or have access to the
1.70      gdt       751: dom0.  This is not intended to be an exhaustive list of VPS providers;
                    752: only a few are mentioned that specifically support NetBSD.
1.28      gdt       753: 
1.52      gdt       754: VPS operators provide varying degrees of access and mechanisms for
                    755: configuration.  The big issue is usually how one controls which kernel
1.111     wiz       756: is booted, because the kernel is nominally in the dom0 file system (to
                    757: which VPS users do not normally have access).  A second issue is how
1.70      gdt       758: to install NetBSD.
1.52      gdt       759: A VPS user may want to compile a kernel for security updates, to run
                    760: npf, run IPsec, or any other reason why someone would want to change
                    761: their kernel.
1.111     wiz       763: One approach is to have an administrative interface to upload a kernel,
1.68      gdt       764: or to select from a prepopulated list.  Other approaches are pygrub
1.59      gdt       765: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
1.111     wiz       766: kernel from the domU file system.  This is closer to a regular physical
1.59      gdt       767: computer, where someone who controls a machine can replace the kernel.
1.52      gdt       768: 
1.74      gdt       769: A second issue is multiple CPUs.  With NetBSD 6, domUs support
                    770: multiple vcpus, and it is typical for VPS providers to enable multiple
                    771: CPUs for NetBSD domUs.
1.180     gdt       773: ## Complexities due to Xen changes
                    775: Xen has many security advisories and people running Xen systems make
                    776: different choices.
                    778: ### stub domains
                    780: Some (Linux only?) dom0 systems use something called "stub domains" to
                    781: isolate qemu from the dom0 system, as a security and reliabilty
                    782: mechanism when running HVM domUs.  Somehow, NetBSD's GENERIC kernel
                    783: ends up using PIO for disks rather than DMA.  Of course, all of this
                    784: is emulated, but emulated PIO is unusably slow.  This problem is not
                    785: currently understood.
                    787: ### Grant tables
                    789: There are multiple versions of using grant tables, and some security
                    790: advisories have suggested disabling some versions.  Some versions of
                    791: NetBSD apparently only use specific versions and this can lead to
                    792: "NetBSD current doesn't run on hosting provider X" situations.
                    794: \todo Explain better.
1.181     gdt       796: ## Boot methods
                    798: ### pvgrub
1.59      gdt       799: 
                    800: pvgrub is a version of grub that uses PV operations instead of BIOS
                    801: calls.  It is booted from the dom0 as the domU kernel, and then reads
1.111     wiz       802: /grub/menu.lst and loads a kernel from the domU file system.
1.59      gdt       803: 
1.70      gdt       804: [Panix]( lets users use pvgrub.  Panix reports
1.71      gdt       805: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
                    806: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
1.70      gdt       807: page](, which describes only
1.74      gdt       808: Linux but should be updated to cover NetBSD :-).
1.70      gdt       809: 
                    810: []( also lets users with pvgrub to boot
                    811: their own kernel.  See then [ NetBSD
1.74      gdt       812: HOWTO](
                    813: (which is in need of updating).
1.59      gdt       814: 
1.70      gdt       815: It appears that [grub's FFS
                    816: code](
                    817: does not support all aspects of modern FFS, but there are also reports
1.72      gdt       818: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
1.70      gdt       819: partition for the kernel with the intent that grub can understand it,
                    820: which leads to /netbsd not being the actual kernel.  One must remember
1.111     wiz       821: to update the special boot partition.
1.59      gdt       822: 
1.181     gdt       823: ### pygrub
1.168     wiki      824: 
                    825: pygrub runs in the dom0 and looks into the domU file system.  This
                    826: implies that the domU must have a kernel in a file system in a format
                    827: known to pygrub.
                    829: pygrub doesn't seem to work to load Linux images under NetBSD dom0,
                    830: and is inherently less secure than pvgrub due to running inside dom0. For both these
                    831: reasons, pygrub should not be used, and is only still present so that
                    832: historical DomU images using it still work.
                    834: As of 2014, pygrub seems to be of mostly historical
                    835: interest. New DomUs should use pvgrub.
1.181     gdt       837: ## Specific Providers
                    839: ### Amazon
1.59      gdt       840: 
1.143     wiki      841: See the [Amazon EC2 page](/amazon_ec2/).

CVSweb for NetBSD wikisrc <> software: FreeBSD-CVSweb