File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.207: download - view: text, annotated - select for diffs
Thu Apr 15 15:23:22 2021 UTC (3 months, 1 week ago) by gdt
Branches: MAIN
CVS tags: HEAD
xen howto: retract hopeful statement about EFI

on rereading the thread, it isn't supported

    1: [[!meta title="Xen Status and HowTo"]]
    2: 
    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.
   10: 
   11: This document provides status on what Xen things work on NetBSD
   12: (upstream documentation might say something works if it works on some
   13: particular Linux system).
   14: 
   15: This document is also a HOWTO that presumes a basic familiarity with
   16: the Xen system architecture, with installing NetBSD on amd64 hardware,
   17: and with installing software from pkgsrc.  See also the [Xen
   18: website](http://www.xenproject.org/).
   19: 
   20: If this document says that something works, and you find that it does
   21: not, it is best to ask on port-xen and if you are correct to file a
   22: PR.
   23: 
   24: [[!toc]]
   25: 
   26: # Overview
   27: 
   28: The basic concept of Xen is that the hypervisor (xenkernel) runs on
   29: the hardware, and runs a privileged domain ("dom0") that can access
   30: disks/networking/etc.  One then runs additional unprivileged domains
   31: (each a "domU"), presumably to do something useful.
   32: 
   33: This HOWTO addresses how to run a NetBSD dom0 (and hence also build
   34: xen itself).  It also addresses how to run domUs in that environment,
   35: and how to deal with having a domU in a Xen environment run by someone
   36: else and/or not running NetBSD.
   37: 
   38: There are many choices one can make; the HOWTO recommends the standard
   39: approach and limits discussion of alternatives in many cases.
   40: 
   41: ## Guest Styles
   42: 
   43: Xen supports different styles of guests.  See
   44: https://wiki.xenproject.org/wiki/Virtualization_Spectrum for a
   45: discussion.
   46: 
   47: This table shows the styles, and if a NetBSD dom0 can run in that
   48: style, if a NetBSD dom0 can sypport that style of guest in a domU, and
   49: if NetBSD as a domU can support that style.
   50: 
   51: [[!table data="""
   52: Style of guest	|dom0 can be?	|dom0 can support?	|domU can be?
   53: PV		|yes		|yes			|yes
   54: HVM		|N/A		|yes			|yes
   55: PVHVM		|N/A		|yes			|current only
   56: PVH		|not yet	|current only		|current only
   57: """]]
   58: 
   59: In PV (paravirtualized) mode, the guest OS does not attempt to access
   60: hardware directly, but instead makes hypercalls to the hypervisor; PV
   61: guests must be specifically coded for Xen.  See
   62: [PV](https://wiki.xen.org/wiki/Paravirtualization_(PV\)).
   63: 
   64: In HVM (Hardware Virtual Machine) mode, no guest modification is
   65: required.  However, hardware support is required, such as VT-x on
   66: Intel CPUs and SVM on AMD CPUs to assist with the processor emulation.
   67: The dom0 runs qemu to emulate hardware other than the processor.  It
   68: is therefore non-sensical to have an HVM dom0, because there is no
   69: underlying system to provide emulation.
   70: 
   71: In PVHVM mode, the guest runs as HVM, but additionally uses PV
   72: drivers for efficiency.  Therefore it is non-sensical for to have a
   73: PVHVM dom0.  See [PV on HVM](https://wiki.xen.org/wiki/PV_on_HVM).
   74: 
   75: There have been two PVH modes: original PVH and PVHv2.  Original PVH
   76: was based on PV mode and is no longer relevant at all.  Therefore
   77: PVHv2 is written as PVH, here and elsewhere.  PVH is basically
   78: lightweight HVM with PV drivers.  A critical feature of it is that
   79: qemu is not needed; the hypervisor can do the emulation that is
   80: required.  Thus, a dom0 can be PVH.  The source code uses PVH and
   81: config files use pvh, but NB that this refers to PVHv2.  See
   82: [PVH(v2)](https://wiki.xenproject.org/wiki/PVH_(v2\)_Domu).
   83: 
   84: At system boot, the dom0 kernel is loaded as a module with Xen as the
   85: kernel.  The dom0 can start one or more domUs.  (Booting is explained
   86: in detail in the dom0 section.)
   87: 
   88: ## CPU Architecture
   89: 
   90: Xen runs on x86_64 hardware (the NetBSD amd64 port).
   91: 
   92: There is a concept of Xen running on ARM, but there are no reports of this working with NetBSD.
   93: 
   94: The dom0 system should be amd64.  (Instructions for i386PAE dom0 have been removed from the HOWTO.)
   95: 
   96: The domU can be i386 PAE or amd64.
   97: i386 PAE at one point was considered as [faster](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html) than amd64.
   98: However, as of 2021 it is normal to use amd64 as the domU architecture, and use of i386 is dwindling.
   99: 
  100: ## Xen Versions
  101: 
  102: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
  103: xenkernel and xentools.  We will refer only to the kernel versions,
  104: but note that both packages must be installed together and must have
  105: matching versions.
  106: 
  107: Versions available in pkgsrc:
  108: 
  109: [[!table data="""
  110: Xen Version	|Package Name	|Xen CPU Support	|EOL'ed By Upstream
  111: 4.11		|xenkernel411	|x86_64			|No
  112: 4.13		|xenkernel413	|x86_64			|No
  113: """]]
  114: 
  115: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
  116: 
  117: Older Xen had a python-based management tool called xm; this has been
  118: replaced by xl.
  119: 
  120: ## NetBSD versions
  121: 
  122: Xen has been supported in NetBSD for a long time, at least since 2005.
  123: Initially Xen was PV only.
  124: 
  125: NetBSD Xen has always supported PV, in both dom0 and domU; for a long
  126: time this was the only way.  NetBSD >=8 as a dom0 supports HVM mode in
  127: domUs.
  128: 
  129: Support for PVHVM and PVH is available only in NetBSD-current; this is
  130: currently somewhat experimental, although PVHVM appears reasonably
  131: solid.
  132: 
  133: NetBSD up to and including NetBSD 9 as a dom0 cannot safely run SMP.
  134: Even if one added "options MULTIPROCESSOR" and configured multiple
  135: vcpus, the kernel is likely to crash because of drivers without
  136: adequate locking.
  137: 
  138: NetBSD-current supports SMP in dom0, and XEN3_DOM0 includes "options
  139: MULTIPROCESSOR".
  140: 
  141: NetBSD (since NetBSD 6), when run as a domU, can run SMP, using
  142: multiple CPUs if provided.  The XEN3_DOMU kernel is built
  143: with "options MULITPROCESSOR".
  144: 
  145: Note that while Xen 4.13 is current, the kernel support is still
  146: called XEN3, because the hypercall interface has not changed
  147: significantly.
  148: 
  149: # Creating a NetBSD dom0
  150: 
  151: In order to install a NetBSD as a dom0, one first installs a normal
  152: NetBSD system, and then pivot the install to a dom0 install by
  153: changing the kernel and boot configuration.
  154: 
  155: NB: As of 2021-04, you must arrange to have the system use BIOS boot,
  156: not EFI boot.
  157: 
  158: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
  159: 512M it worked reliably.  This does not make sense, but if you see
  160: "not ELF" after Xen boots, try increasing dom0 RAM.
  161: 
  162: ## Installation of NetBSD
  163: 
  164: [Install NetBSD/amd64](/guide/inst/) just as you would if you were not
  165: using Xen.  Therefore, use the most recent release, or a build from
  166: the most recent stable branch.  Alternatively, use -current, being
  167: mindful of all the usual caveats of lower stability of current, and
  168: likely a bit more so.  Think about how you will provide storage for
  169: disk images.
  170: 
  171: ## Installation of Xen
  172: 
  173: ### Building Xen
  174: 
  175: Use the most recent version of Xen in pkgsrc, unless the DESCR says
  176: that it is not suitable.  Therefore, choose 4.13.  In the dom0,
  177: install xenkernel413 and xentools413 from pkgsrc.
  178: 
  179: Once this is done, copy the Xen kernel from where pkgsrc puts it to
  180: where the boot process will be able to find it:
  181: 
  182: [[!template id=programlisting text="""
  183: # cp -p /usr/pkg/xen413-kernel/xen.gz /
  184: """]]
  185: 
  186: Then, place a NetBSD XEN3_DOM0 kernel in the `/` directory. Such
  187: kernel can either be taken from a local release build.sh run, compiled
  188: manually, or downloaded from the NetBSD FTP, for example at:
  189: 
  190: [[!template id=programlisting text="""
  191: ftp.netbsd.org/pub/NetBSD/NetBSD-9.1/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
  192: """]]
  193: 
  194: ### Configuring booting
  195: 
  196: Read boot.cfg(8) carefully.  Add lines to /boot.cfg to boot Xen,
  197: adjusting for your root filesystem:
  198: 
  199: [[!template id=filecontent name="/boot.cfg" text="""
  200: menu=Xen:load /netbsd-XEN3_DOM0.gz root=wd0a console=pc;multiboot /xen.gz dom0_mem=512M
  201: menu=Xen single user:load /netbsd-XEN3_DOM0.gz root=wd0a console=pc -s;multiboot /xen.gz dom0_mem=512M
  202: """]]
  203: 
  204: This specifies that the dom0 should have 512MB of ram, leaving the rest
  205: to be allocated for domUs.
  206: 
  207: NB: This says add, not replace, so that you will be able to more
  208: easily boot a NetBSD kernel without Xen.  Once Xen boots ok, you may
  209: want to set it as default.  It is highly likely that you will have
  210: trouble at some point, and keeping an up-to-date GENERIC for use in
  211: fixing problems is the standard prudent approach.
  212: 
  213: \todo Explain why rndseed is not set with Xen as part of the dom0
  214: subconfiguration.
  215: 
  216: Note that you are likely to have to set root= because the boot device
  217: from /boot is not passed via Xen to the dom0 kernel.  With one disk,
  218: it will work, but e.g. plugging in USB disk to a machine with root on
  219: wd0a causes boot to fail.
  220: 
  221: Beware that userconf statements must be attached to the dom0 load, and
  222: may not be at top-level, because then they would try to configure the
  223: hypervisor, if there is a way to pass them via multiboot.  It appears
  224: that adding `userconf=pckbc` to `/boot.cfg` causes Xen to crash very
  225: early with a heap overflow.
  226: 
  227: ### Console selection
  228: 
  229: See boot_console(8).  Understand that you should start from a place of
  230: having console setup correct for booting GENERIC before trying to
  231: configure Xen.
  232: 
  233: Generally for GENERIC, one sets the console in bootxx_ffsv1 or
  234: equivalent, and this is passed on to /boot (where one typically does
  235: not set the console).  This configuration of bootxx_ffsv1 should also
  236: be in place for Xen systems, to allow seeing messages from /boot and
  237: use of a keyboard to select a line from the menu.  And, one should
  238: have a working boot path to GENERIC for rescue situations.
  239: 
  240: With GENERIC, the boot options are passed on to /netbsd, but there is
  241: currently no mechanism to pass these via multiboot to the hypervisor.
  242: Thus, in addition to configuring the console in the boot blocks, one
  243: must also configure it for Xen.
  244: 
  245: By default, the hypervisor (Xen itself) will use some sort of vga
  246: device as the console, much like GENERIC uses by default.  The vga
  247: console is relinquished at the conclusion of hypervisor boot, before
  248: the dom0 is started.  Xen when using a vga console does not process
  249: console input.
  250: 
  251: The hypervisor can be configured to use a serial port console, e.g.
  252: [[!template id=filecontent name="/boot.cfg" text="""
  253: menu=Xen:load /netbsd-XEN3_DOM0.gz console=com0;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
  254: """]]
  255: This example uses the first serial port (Xen counts from 1; this is
  256: what NetBSD would call com0), and sets speed and parity.  (The dom0 is
  257: then configured to use the same serial port in this example.)
  258: 
  259: With the hypervisor configured for a serial console, it can get input,
  260: and there is a notion of passing this input to the dom0.  \todo
  261: Explain why, if Xen has a serial console, the dom0 console is
  262: typically also configured to open that same serial port, instead of
  263: getting the passthrough input via the xen console.
  264: 
  265: One also configures the console for the dom0.  While one might expect
  266: console=pc to be default, following behavior of GENERIC, a hasty read
  267: of the code suggests there is no default and booting without a
  268: selected console might lead to a panic.  Also, there is merit in
  269: explicit configuration.  Therefore the standard approach is to place
  270: console=pc as part of the load statement for the dom0 kernel, or
  271: alternatively console=com0.
  272: 
  273: The NetBSD dom0 kernel will attach xencons(4) (the man page does not
  274: exist), but this is not used as a console.  It is used to obtain the
  275: messages from the hypervisor's console; run `xl dmesg` to see them.
  276: 
  277: ### Tuning
  278: 
  279: In an attempt to add performance, one can also add `dom0_max_vcpus=1
  280: dom0_vcpus_pin`, to force only one vcpu to be provided (since NetBSD
  281: dom0 can't use more) and to pin that vcpu to a physical CPU. Xen has
  282: [many boot
  283: options](http://xenbits.xenproject.org/docs/4.13-testing/misc/xen-command-line.html),
  284: and other than dom0 memory and max_vcpus, they are generally not
  285: necessary.
  286: 
  287: \todo Revisit this advice with current.
  288: \todo Explain if anyone has ever actually measured that this helps.
  289: 
  290: ### rc.conf
  291: 
  292: Ensure that the boot scripts installed in
  293: `/usr/pkg/share/examples/rc.d` are in `/etc/rc.d`, either because you
  294: have `PKG_RCD_SCRIPTS=yes`, or manually.  (This is not special to Xen,
  295: but a normal part of pkgsrc usage.)
  296: 
  297: Set `xencommons=YES` in rc.conf:
  298: 
  299: [[!template id=filecontent name="/etc/rc.conf" text="""
  300: xencommons=YES
  301: """]]
  302: 
  303: \todo Recommend for/against xen-watchdog.
  304: 
  305: ### Testing
  306: 
  307: Now, reboot so that you are running a DOM0 kernel under Xen, rather
  308: than GENERIC without Xen.
  309: 
  310: Once the reboot is done, use `xl` to inspect Xen's boot messages,
  311: available resources, and running domains.  For example:
  312: 
  313: [[!template id=programlisting text="""
  314: # xl dmesg
  315: ... xen's boot info ...
  316: # xl info
  317: ... available memory, etc ...
  318: # xl list
  319: Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  320: Domain-0           0       64    0  r----     58.1
  321: """]]
  322: 
  323: Xen logs will be in /var/log/xen.
  324: 
  325: ### Issues with xencommons
  326: 
  327: `xencommons` starts `xenstored`, which stores data on behalf of dom0 and
  328: domUs.  It does not currently work to stop and start xenstored.
  329: Certainly all domUs should be shutdown first, following the sort order
  330: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
  331: and is not notified when xenstored exits, leading to not recreating
  332: the state when the new xenstored starts.  Until there's a mechanism to
  333: make this work, one should not expect to be able to restart xenstored
  334: (and thus xencommons).  There is currently no reason to expect that
  335: this will get fixed any time soon.
  336: \todo Confirm if this is still true in 2020.
  337: 
  338: ## Xen-specific NetBSD issues
  339: 
  340: There are (at least) two additional things different about NetBSD as a
  341: dom0 kernel compared to hardware.
  342: 
  343: One is that through NetBSD 9 the module ABI is different because some
  344: of the #defines change, so there are separate sets of modules in
  345: /stand.  (Further, zfs in Xen is troubled because of differing
  346: MAXPHYS; see the zfs howto for more.)  In NetBSD-current, there is
  347: only one set of modules.
  348: 
  349: The other difference is that XEN3_DOM0 does not have exactly the same
  350: options as GENERIC.  While this is roughly agreed to be in large part
  351: a bug, users should be aware of this and can simply add missing config
  352: items if desired.
  353: 
  354: Finally, there have been occasional reports of trouble with X11
  355: servers in NetBSD as a dom0.  Some hardware support is intentionally
  356: disabled in XEN3_DOM0.
  357: 
  358: ## Updating Xen in a dom0
  359: 
  360: Note the previous advice to maintain a working and tested boot config
  361: into GENERIC without Xen.
  362: 
  363: Updating Xen in a dom0 consists of updating the xnekernel and xentools
  364: packages, along with copying the xen.gz into place, and of course
  365: rebooting.
  366: 
  367: If updating along a Xen minor version, e.g. from 4.13.1 to 4.13.2, or
  368: from 4.13.2nb1 to 4.13.2nb3, it is very likely that this can be done
  369: on a running system.  The point is that the xentools programs will be
  370: replaced, and you will be using "xl" from the new installation to talk
  371: to the older programs which are still running.  Problems from this
  372: update path should be reported.
  373: 
  374: For added safety, shutdown all domUs before updating, to remove the
  375: need for new xl to talk to old xenstored.  Note that Xen does not
  376: guarantee stability of internal ABIs.
  377: 
  378: If updating across Xen minor versions, e.g. from 4.11 to 4.13, the
  379: likelihood of trouble is increased.  Therefore, 'make replace' of
  380: xentools on a dom0 with running domUs is not recommended.  A shutdown
  381: on all domUs before replacing xentools is likely sufficient.  A safer
  382: appraoch is to boot into GENERIC to replace the packages, as then no
  383: Xen code will be running.  Single user is another option.
  384: 
  385: ## Updating NetBSD in a dom0
  386: 
  387: This is just like updating NetBSD on bare hardware, assuming the new
  388: version supports the version of Xen you are running.  Generally, one
  389: replaces the kernel and reboots, and then overlays userland binaries
  390: and adjusts `/etc`.
  391: 
  392: Note that one should update both the non-Xen kernel typically used for
  393: rescue purposes, as well as the DOM0 kernel used with Xen.
  394: 
  395: ## anita (for testing NetBSD)
  396: 
  397: With a NetBSD dom0, even without any domUs, one can run anita (see
  398: pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as root,
  399: because anita must create a domU):
  400: 
  401: [[!template id=programlisting text="""
  402: anita --vmm=xl test file:///usr/obj/i386/
  403: """]]
  404: 
  405: # Unprivileged domains (domU)
  406: 
  407: This section describes general concepts about domUs.  It does not
  408: address specific domU operating systems or how to install them.  The
  409: config files for domUs are typically in `/usr/pkg/etc/xen`, and are
  410: typically named so that the file name, domU name and the domU's host
  411: name match.
  412: 
  413: The domU is provided with CPU and memory by Xen, configured by the
  414: dom0.  The domU is provided with disk and network by the dom0,
  415: mediated by Xen, and configured in the dom0.
  416: 
  417: Entropy in domUs can be an issue; physical disks and network are on
  418: the dom0.  NetBSD's /dev/random system works, but is often challenged.
  419: 
  420: ## Config files
  421: 
  422: See /usr/pkg/share/examples/xen/xlexample* for a very small number of
  423: examples for running GNU/Linux.
  424: 
  425: The following is an example minimal domain configuration file.  The
  426: domU serves as a network file server.
  427: 
  428: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
  429: name = "domU-id"
  430: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
  431: memory = 1024
  432: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
  433: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
  434:          'file:/n0/xen/foo-wd1,0x1,w' ]
  435: """]]
  436: 
  437: The domain will have name given in the `name` setting.  The kernel has the
  438: host/domU name in it, so that on the dom0 one can update the various
  439: domUs independently.  The `vif` line causes an interface to be provided,
  440: with a specific mac address (do not reuse MAC addresses!), in bridge
  441: mode.  Two disks are provided, and they are both writable; the bits
  442: are stored in files and Xen attaches them to a vnd(4) device in the
  443: dom0 on domain creation.  The system treats xbd0 as the boot device
  444: without needing explicit configuration.
  445: 
  446: There is not a type line; that implicitly defines a pv domU.
  447: Otherwise, one sets type to the lower-case version of the domU type in
  448: the table above; see later sections.
  449: 
  450: By convention, domain config files are kept in `/usr/pkg/etc/xen`.  Note
  451: that "xl create" takes the name of a config file, while other commands
  452: take the name of a domain.
  453: 
  454: Examples of commands:
  455: 
  456: [[!template id=programlisting text="""
  457: xl create /usr/pkg/etc/xen/foo
  458: xl console domU-id
  459: xl create -c /usr/pkg/etc/xen/foo
  460: xl shutdown domU-id
  461: xl list
  462: """]]
  463: 
  464: Typing `^]` will exit the console session.  Shutting down a domain is
  465: equivalent to pushing the power button; a NetBSD domU will receive a
  466: power-press event and do a clean shutdown.  Shutting down the dom0
  467: will trigger controlled shutdowns of all configured domUs.
  468: 
  469: ## CPU and memory
  470: 
  471: A domain is provided with some number of vcpus; any domain can have up
  472: to the number of CPUs seen by the hypervisor. For a domU, it is
  473: controlled from the config file by the "vcpus = N" directive.  It is
  474: normal to overcommit vcpus; a 4-core machine machine might well provide 4
  475: vcpus to each domU.  One might also configure fewer vcpus for a domU.
  476: 
  477: A domain is provided with memory; this is controlled in the config
  478: file by "memory = N" (in megabytes).  In the straightforward case, the
  479: sum of the the memory allocated to the dom0 and all domUs must be less
  480: than the available memory.
  481: 
  482: ## Balloon driver
  483: 
  484: Xen provides a `balloon` driver, which can be used to let domains use
  485: more memory temporarily.
  486: 
  487: \todo Explain how to set up a aystem to use the balloon scheme in a
  488: useful manner.
  489: 
  490: ## Virtual disks
  491: 
  492: In domU config files, the disks are defined as a sequence of 3-tuples:
  493: 
  494:  * The first element is "method:/path/to/disk". Common methods are
  495:    "file:" for a file-backed vnd, and "phy:" for something that is already
  496:    a device, such as an LVM logical volume.
  497: 
  498:  * The second element is an artifact of how virtual disks are passed to
  499:    Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
  500:    are given a device name to associate with the disk, and values like
  501:    "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
  502:    as xbd0, the second as xbd1, and so on.  However, xl demands a
  503:    second argument.  The name given is converted to a major/minor by
  504:    calling stat(2) on the name in /dev and this is passed to the domU.
  505:    In the general case, the dom0 and domU can be different operating
  506:    systems, and it is an unwarranted assumption that they have consistent
  507:    numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
  508:    as both dom0 and domU, using values of 0x0 for the first disk and 0x1
  509:    for the second works fine and avoids this issue.  For a GNU/Linux
  510:    guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
  511:    /dev/hda1.
  512: 
  513:  * The third element is "w" for writable disks, and "r" for read-only
  514:    disks.
  515: 
  516: Example:
  517: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
  518: disk = [ 'file:/n0/xen/foo-wd0,0x0,w' ]
  519: """]]
  520: 
  521: Note that NetBSD by default creates only vnd[0123].  If you need more
  522: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
  523: dom0.
  524: 
  525: ## Virtual Networking
  526: 
  527: Xen provides virtual Ethernets, each of which connects the dom0 and a
  528: domU.  For each virtual network, there is an interface "xvifN.M" in
  529: the dom0, and a matching interface xennetM (NetBSD name) in domU index N.
  530: The interfaces behave as if there is an Ethernet with two
  531: adapters connected.  From this primitive, one can construct various
  532: configurations.  We focus on two common and useful cases for which
  533: there are existing scripts: bridging and NAT.
  534: 
  535: With bridging (in the example above), the domU perceives itself to be
  536: on the same network as the dom0.  For server virtualization, this is
  537: usually best.  Bridging is accomplished by creating a bridge(4) device
  538: and adding the dom0's physical interface and the various xvifN.0
  539: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
  540: config file.  The bridge must be set up already in the dom0; an
  541: example /etc/ifconfig.bridge0 is:
  542: 
  543: [[!template id=filecontent name="/etc/ifconfig.bridge0" text="""
  544: create
  545: up
  546: !brconfig bridge0 add wm0
  547: """]]
  548: 
  549: With NAT, the domU perceives itself to be behind a NAT running on the
  550: dom0.  This is often appropriate when running Xen on a workstation.
  551: TODO: NAT appears to be configured by "vif = [ '' ]".
  552: 
  553: The MAC address specified is the one used for the interface in the new
  554: domain.  The interface in dom0 will use this address XOR'd with
  555: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
  556: 
  557: ## Starting domains automatically
  558: 
  559: To start domains `domU-netbsd` and `domU-linux` at boot and shut them
  560: down cleanly on dom0 shutdown, add the following in rc.conf:
  561: 
  562: [[!template id=filecontent name="/etc/rc.conf" text="""
  563: xendomains="domU-netbsd domU-linux"
  564: """]]
  565: 
  566: # domU setup for specific systems
  567: 
  568: Creating domUs is almost entirely independent of operating system.  We
  569: have already presented the basics of config files in the previous system.
  570: 
  571: Of course, this section presumes that you have a working dom0.
  572: 
  573: ## Creating a NetBSD PV domU
  574: 
  575: See the earlier config file, and adjust memory.  Decide on how much
  576: storage you will provide, and prepare it (file or LVM).
  577: 
  578: While the kernel will be obtained from the dom0 file system, the same
  579: file should be present in the domU as /netbsd so that tools like
  580: savecore(8) can work.   (This is helpful but not necessary.)
  581: 
  582: The kernel must be specifically built for Xen, to use PV interfacesas
  583: a domU.  NetBSD release builds provide the following kernels:
  584: 
  585:         i386 XEN3PAE_DOMU
  586:         amd64 XEN3_DOMU
  587: 
  588: This will boot NetBSD, but this is not that useful if the disk is
  589: empty.  One approach is to unpack sets onto the disk outside of Xen
  590: (by mounting it, just as you would prepare a physical disk for a
  591: system you can't run the installer on).
  592: 
  593: A second approach is to run an INSTALL kernel, which has a miniroot
  594: and can load sets from the network.  To do this, copy the INSTALL
  595: kernel to / and change the kernel line in the config file to:
  596: 
  597:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  598: 
  599: Then, start the domain as "xl create -c configfile".
  600: 
  601: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
  602: line should be used in the config file.
  603: 
  604:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  605: 
  606: After booting the domain, the option to install via CDROM may be
  607: selected.  The CDROM device should be changed to `xbd1d`.
  608: 
  609: Once done installing, "halt -p" the new domain (don't reboot or halt:
  610: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
  611: config file), switch the config file back to the XEN3_DOMU kernel,
  612: and start the new domain again. Now it should be able to use "root on
  613: xbd0a" and you should have a functional NetBSD domU.
  614: 
  615: TODO: check if this is still accurate.
  616: When the new domain is booting you'll see some warnings about *wscons*
  617: and the pseudo-terminals. These can be fixed by editing the files
  618: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  619: `/etc/ttys`, except *console*, like this:
  620: 
  621:     console "/usr/libexec/getty Pc"         vt100   on secure
  622:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  623:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  624:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  625:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  626: 
  627: Finally, all screens must be commented out from `/etc/wscons.conf`.
  628: 
  629: One should also run `powerd` in a domU, but this should not need
  630: configuring.  With powerd, the domain will run a controlled shutdown
  631: if `xl shutdown -R` or `xl shutdown -H` is used on the dom0, via
  632: receiving a synthetic `power button pressed` signal.  In 9 and
  633: current, `powerd` is run by default under Xen kernels (or if ACPI is
  634: present), and it can be added to rc.conf if not.
  635: 
  636: It is not strictly necessary to have a kernel (as /netbsd) in the domU
  637: file system.  However, various programs (e.g. netstat) will use that
  638: kernel to look up symbols to read from kernel virtual memory.  If
  639: /netbsd is not the running kernel, those lookups will fail.  (This is
  640: not really a Xen-specific issue, but because the domU kernel is
  641: obtained from the dom0, it is far more likely to be out of sync or
  642: missing with Xen.)
  643: 
  644: Note that NetBSD by default creates only xbd[0123].  If you need more
  645: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
  646: 
  647: ## Creating a Linux PV domU
  648: 
  649: Creating unprivileged Linux domains isn't much different from
  650: unprivileged NetBSD domains, but there are some details to know.
  651: 
  652: First, the second parameter passed to the disk declaration (the '0x1' in
  653: the example below)
  654: 
  655:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  656: 
  657: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  658: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
  659: So, hda1 which has major 3 and minor 1 on a Linux system will have
  660: device number 0x301.  Alternatively, devices names can be used (hda,
  661: hdb, ...)  as xentools has a table to map these names to devices
  662: numbers.  To export a partition to a Linux guest we can use:
  663: 
  664:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
  665:         root = "/dev/hda1 ro"
  666: 
  667: and it will appear as /dev/hda on the Linux system, and be used as root
  668: partition.
  669: 
  670: To install the Linux system on the partition to be exported to the
  671: guest domain, the following method can be used: install
  672: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
  673: that will be the root partition of your Linux domain, and mount it.
  674: Then copy the files from a working Linux system, make adjustments in
  675: `/etc` (fstab, network config).  It should also be possible to extract
  676: binary packages such as .rpm or .deb directly to the mounted partition
  677: using the appropriate tool, possibly running under NetBSD's Linux
  678: emulation.  Once the file system has been populated, umount it.  If
  679: desirable, the file system can be converted to ext3 using tune2fs -j.
  680: It should now be possible to boot the Linux guest domain, using one of
  681: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
  682: 
  683: To get the Linux console right, you need to add:
  684: 
  685:     extra = "xencons=tty1"
  686: 
  687: to your configuration since not all Linux distributions auto-attach a
  688: tty to the xen console.
  689: 
  690: ## Creating a NetBSD HVM domU
  691: 
  692: Use type='hvm', probably.  Use a GENERIC kernel within the disk image.
  693: 
  694: ## Creating a NetBSD PVH domU
  695: 
  696: This only works with a current kernel in the domU.
  697: 
  698: Use type='pvh'.  Probably, use a GENERIC kernel within the disk image,
  699: which in current has PV support.
  700: 
  701: \todo Verify.
  702: 
  703: \todo Verify if one can have current PVH domU on a 9 dom0.
  704: 
  705: ## Creating a Solaris domU
  706: 
  707: See possibly outdated
  708: [Solaris domU instructions](/ports/xen/howto-solaris/).
  709: 
  710: ## PCI passthrough: Using PCI devices in guest domains
  711: 
  712: NB: PCI passthrough only works on some Xen versions and as of 2020 it
  713: is not clear that it works on any version in pkgsrc.  \todo Reports
  714: confirming or denying this notion should be sent to port-xen@.
  715: 
  716: The dom0 can give other domains access to selected PCI
  717: devices. This can allow, for example, a non-privileged domain to have
  718: access to a physical network interface or disk controller.  However,
  719: keep in mind that giving a domain access to a PCI device most likely
  720: will give the domain read/write access to the whole physical memory,
  721: as PCs don't have an IOMMU to restrict memory access to DMA-capable
  722: device.  Also, it's not possible to export ISA devices to non-dom0
  723: domains, which means that the primary VGA adapter can't be exported.
  724: A guest domain trying to access the VGA registers will panic.
  725: 
  726: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
  727: not been ported to later versions at this time.
  728: 
  729: For a PCI device to be exported to a domU, is has to be attached to
  730: the "pciback" driver in dom0.  Devices passed to the dom0 via the
  731: pciback.hide boot parameter will attach to "pciback" instead of the
  732: usual driver.  The list of devices is specified as "(bus:dev.func)",
  733: where bus and dev are 2-digit hexadecimal numbers, and func a
  734: single-digit number:
  735: 
  736:         pciback.hide=(00:0a.0)(00:06.0)
  737: 
  738: pciback devices should show up in the dom0's boot messages, and the
  739: devices should be listed in the `/kern/xen/pci` directory.
  740: 
  741: PCI devices to be exported to a domU are listed in the "pci" array of
  742: the domU's config file, with the format "0000:bus:dev.func".
  743: 
  744:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  745: 
  746: In the domU an "xpci" device will show up, to which one or more pci
  747: buses will attach.  Then the PCI drivers will attach to PCI buses as
  748: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
  749: or any PCI drivers built in by default; you have to build your own
  750: kernel to use PCI devices in a domU.  Here's a kernel config example;
  751: note that only the "xpci" lines are unusual.
  752: 
  753:         include         "arch/i386/conf/XEN3_DOMU"
  754: 
  755:         # Add support for PCI buses to the XEN3_DOMU kernel
  756:         xpci* at xenbus ?
  757:         pci* at xpci ?
  758: 
  759:         # PCI USB controllers
  760:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  761: 
  762:         # USB bus support
  763:         usb*    at uhci?
  764: 
  765:         # USB Hubs
  766:         uhub*   at usb?
  767:         uhub*   at uhub? port ? configuration ? interface ?
  768: 
  769:         # USB Mass Storage
  770:         umass*  at uhub? port ? configuration ? interface ?
  771:         wd*     at umass?
  772:         # SCSI controllers
  773:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  774: 
  775:         # SCSI bus support (for both ahc and umass)
  776:         scsibus* at scsi?
  777: 
  778:         # SCSI devices
  779:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
  780:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  781: 
  782: 
  783: # Miscellaneous Information
  784: 
  785: ## Nesting under Linux KVM
  786: 
  787: It is possible to run Xen and a NetBSD dom0 under Linux KVM.  One
  788: can enable virtio in the dom0 for greater speed.
  789: 
  790: ## Nesting under qemu
  791: 
  792: It is possible to run Xen and a NetBSD dom0 under qemu on NetBSD, and
  793: also with nvmm.
  794: \todo Check this.
  795: 
  796: ## Other nesting
  797: 
  798: In theory, any full emulation should be able to run Xen and a NetBSD
  799: dom0.  The HOWTO does not currently have information about Xen XVM
  800: mode, Virtualbox, etc.
  801: 
  802: ## NetBSD 5 as domU
  803: 
  804: [NetBSD 5 is known to panic.](http://mail-index.netbsd.org/port-xen/2018/04/17/msg009181.html)
  805: (However, NetBSD 5 systems should be updated to a supported version.)
  806: 
  807: # NetBSD as a domU in a VPS
  808: 
  809: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
  810: hardware.  This section explains how to deal with Xen in a domU as a
  811: virtual private server where you do not control or have access to the
  812: dom0.  This is not intended to be an exhaustive list of VPS providers;
  813: only a few are mentioned that specifically support NetBSD.
  814: 
  815: VPS operators provide varying degrees of access and mechanisms for
  816: configuration.  The big issue is usually how one controls which kernel
  817: is booted, because the kernel is nominally in the dom0 file system (to
  818: which VPS users do not normally have access).  A second issue is how
  819: to install NetBSD.
  820: A VPS user may want to compile a kernel for security updates, to run
  821: npf, run IPsec, or any other reason why someone would want to change
  822: their kernel.
  823: 
  824: One approach is to have an administrative interface to upload a kernel,
  825: or to select from a prepopulated list.  Other approaches are pygrub
  826: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
  827: kernel from the domU file system.  This is closer to a regular physical
  828: computer, where someone who controls a machine can replace the kernel.
  829: 
  830: A second issue is multiple CPUs.  With NetBSD 6, domUs support
  831: multiple vcpus, and it is typical for VPS providers to enable multiple
  832: CPUs for NetBSD domUs.
  833: 
  834: ## Complexities due to Xen changes
  835: 
  836: Xen has many security advisories and people running Xen systems make
  837: different choices.
  838: 
  839: ### stub domains
  840: 
  841: Some (Linux) dom0 systems use something called "stub domains" to
  842: isolate qemu from the dom0 system, as a security and reliabilty
  843: mechanism when running HVM domUs.  Somehow, NetBSD's GENERIC kernel
  844: ends up using PIO for disks rather than DMA.  Of course, all of this
  845: is emulated, but emulated PIO is unusably slow.  This problem is not
  846: currently understood.
  847: 
  848: ### Grant tables
  849: 
  850: There are multiple versions of using grant tables, and some security
  851: advisories have suggested disabling some versions.  NetBSD through 9
  852: uses version 1 and NetBSD-current uses version 2.  This can lead to
  853: "NetBSD current doesn't run on hosting provider X" situations.
  854: 
  855: \todo Explain better.
  856: 
  857: ## Boot methods
  858: 
  859: ### pvgrub
  860: 
  861: pvgrub is a version of grub that uses PV operations instead of BIOS
  862: calls.  It is booted from the dom0 as the domU kernel, and then reads
  863: /grub/menu.lst and loads a kernel from the domU file system.
  864: 
  865: It appears that [grub's FFS
  866: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
  867: does not support all aspects of modern FFS, but there are also reports
  868: that FFSv2 works fine.
  869: 
  870: ### pygrub
  871: 
  872: As of 2014, pygrub seems to be of mostly historical interest.  As of
  873: 2021, the section should perhaps be outright deleted.
  874: 
  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
  877: known to pygrub.
  878: 
  879: pygrub doesn't seem to work to load Linux images under NetBSD dom0,
  880: and is inherently less secure than pvgrub due to running inside
  881: dom0. For both these reasons, pygrub should not be used, and is only
  882: still present so that historical DomU images using it still work.
  883: 
  884: ## Specific Providers
  885: 
  886: The intent is to list providers only if they document support for
  887: running NetBSD, and to point to their resources briefly.
  888: 
  889: ### panix.com
  890: 
  891: [Panix](http://www.panix.com/) provides NetBSD as an OS option.  See
  892: https://www.panix.com/v-colo/nupgrade.html for some information.
  893: Users can use pvgrub.  Panix reports that pvgrub works with FFsv2 with
  894: 16K/2K and 32K/4K block/frag sizes (and hence with defaults from
  895: "newfs -O 2").  See [Panix's pvgrub
  896: page](http://www.panix.com/v-colo/grub.html) which describes how to
  897: boot NetBSD.
  898: 
  899: ### prgmr.com
  900: 
  901: [prgmr.com](http://prgmr.com/) provides released versions of
  902: NetBSD/amd64 as installation options.  Users can use pvgrub to boot
  903: their own kernel, and a small FAT32 /boot is encouraged.  See the
  904: [prgmr.com NetBSD
  905: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
  906: (which is in need of updating).
  907: 
  908: ### Amazon
  909: 
  910: See the [Amazon EC2 page](/amazon_ec2/).

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