Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.181

1.144     maxv        1: [[!meta title="Xen HowTo"]]
                      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: 
1.179     gdt        11: Xen supports different styles of guests; see [PV on HVM](https://wiki.xen.org/wiki/PV_on_HVM) and [PVH(v2)](https://wiki.xenproject.org/wiki/PVH_(v2\)_Domu) for upstream documentation.
1.149     maxv       12: 
                     13: [[!table data="""
                     14: Style of guest |Supported by NetBSD
1.150     maxv       15: PV             |Yes (dom0, domU)
                     16: HVM            |Yes (domU)
1.171     gdt        17: PVHVM          |current-only (domU)
1.180     gdt        18: PVH            |current-only (domU, dom0 not yet)
1.149     maxv       19: """]]
                     20: 
                     21: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
                     22: hardware directly, but instead makes hypercalls to the hypervisor; PV
1.170     gdt        23: guests must be specifically coded for Xen.
                     24: 
                     25: In HVM mode, no guest modification is required; however, hardware
                     26: support is required, such as VT-x on Intel CPUs and SVM on AMD CPUs.
1.177     gdt        27: The dom0 runs qemu to emulate hardware.
1.170     gdt        28: 
1.176     gdt        29: In PVHVM mode, the guest runs as HVM, but additionally can use PV
                     30: drivers for efficiency.
1.170     gdt        31: 
1.180     gdt        32: There have been two PVH modes: original PVH and PVHv2.  Original PVH
                     33: was based on PV mode and is no longer relevant at all.  PVHv2 is
                     34: basically lightweight HVM with PV drivers.  A critical feature of it
                     35: is that qemu is not needed; the hypervisor can do the emulation that
                     36: is required.  Thus, a dom0 can be PVHv2.
                     37: 
                     38: The source code uses PVH and config files use pvh; this refers to PVHv2.
1.149     maxv       39: 
1.27      jnemeth    40: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12      gdt        41: The dom0 can start one or more domUs.  (Booting is explained in detail
                     42: in the dom0 section.)
                     43: 
                     44: This HOWTO presumes a basic familiarity with the Xen system
1.123     gdt        45: architecture, with installing NetBSD on i386/amd64 hardware, and with
                     46: installing software from pkgsrc.  See also the [Xen
                     47: website](http://www.xenproject.org/).
1.1       mspo       48: 
1.160     maxv       49: [[!toc]]
1.150     maxv       50: 
1.174     gdt        51: # Versions and Support
1.15      gdt        52: 
1.111     wiz        53: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
1.15      gdt        54: xenkernel and xentools.  We will refer only to the kernel versions,
                     55: but note that both packages must be installed together and must have
                     56: matching versions.
                     57: 
1.145     maxv       58: Versions available in pkgsrc:
1.85      gdt        59: 
1.145     maxv       60: [[!table data="""
1.166     gdt        61: Xen Version    |Package Name   |Xen CPU Support        |xm?    |EOL'ed By Upstream
1.167     gdt        62: 4.11           |xenkernel411   |x86_64                 |       |No
1.176     gdt        63: 4.13           |xenkernel413   |x86_64                 |       |No
1.145     maxv       64: """]]
1.113     gdt        65: 
1.96      gdt        66: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
                     67: 
1.157     maxv       68: Multiprocessor (SMP) support in NetBSD differs depending on the domain:
1.145     maxv       69: 
1.157     maxv       70: [[!table data="""
                     71: Domain         |Supports SMP
                     72: dom0           |No
                     73: domU           |Yes
                     74: """]]
1.15      gdt        75: 
1.167     gdt        76: Note: NetBSD support is called XEN3.  However, it does support Xen 4,
1.147     maxv       77: because the hypercall interface has remained identical.
                     78: 
1.166     gdt        79: Older Xen had a python-based management tool called xm, now replaced
1.176     gdt        80: by xl.
1.166     gdt        81: 
1.181   ! gdt        82: ## Architecture
1.18      gdt        83: 
1.167     gdt        84: Xen 4.5 and later runs on x86_64 hardware (the NetBSD amd64 port).
1.176     gdt        85: There is a concept of Xen running on ARM, but there are no reports of this working with NetBSD.
1.157     maxv       86: 
1.176     gdt        87: The dom0 system should be amd64.  (Instructions for i386PAE dom0 have been removed from the HOWTO.)
1.157     maxv       88: 
1.176     gdt        89: The domU can be i386PAE or amd64.
                     90: i386PAE at one point was considered as [faster](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html) than amd64.
1.29      gdt        91: 
1.166     gdt        92: # Creating a dom0
1.15      gdt        93: 
1.158     maxv       94: In order to install a NetBSD as a dom0, one must first install a normal
                     95: NetBSD system, and then pivot the install to a dom0 install by changing
                     96: the kernel and boot configuration.
1.15      gdt        97: 
1.142     gdt        98: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
                     99: 512M it worked reliably.  This does not make sense, but if you see
                    100: "not ELF" after Xen boots, try increasing dom0 RAM.
                    101: 
1.181   ! gdt       102: ## Installation of NetBSD
1.13      gdt       103: 
1.157     maxv      104: [Install NetBSD/amd64](/guide/inst/)
1.19      gdt       105: just as you would if you were not using Xen.
1.1       mspo      106: 
1.181   ! gdt       107: ## Installation of Xen
1.1       mspo      108: 
1.178     gdt       109: We will consider that you chose to use Xen 4.13, with NetBSD/amd64 as
1.155     maxv      110: dom0. In the dom0, install xenkernel48 and xentools48 from pkgsrc.
                    111: 
                    112: Once this is done, install the Xen kernel itself:
                    113: 
                    114: [[!template id=programlisting text="""
                    115: # cp /usr/pkg/xen48-kernel/xen.gz /
                    116: """]]
                    117: 
1.159     maxv      118: Then, place a NetBSD XEN3_DOM0 kernel in the `/` directory. Such kernel
                    119: can either be compiled manually, or downloaded from the NetBSD FTP, for
                    120: example at:
                    121: 
                    122: [[!template id=programlisting text="""
                    123: ftp.netbsd.org/pub/NetBSD/NetBSD-8.0/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
                    124: """]]
1.20      gdt       125: 
1.155     maxv      126: Add a line to /boot.cfg to boot Xen:
1.20      gdt       127: 
1.157     maxv      128: [[!template id=filecontent name="/boot.cfg" text="""
1.152     maxv      129: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
                    130: """]]
1.20      gdt       131: 
1.159     maxv      132: This specifies that the dom0 should have 512MB of ram, leaving the rest
                    133: to be allocated for domUs.  To use a serial console, use:
1.77      gdt       134: 
1.157     maxv      135: [[!template id=filecontent name="/boot.cfg" text="""
1.152     maxv      136: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
                    137: """]]
1.77      gdt       138: 
                    139: which will use the first serial port for Xen (which counts starting
1.139     gson      140: from 1, unlike NetBSD which counts starting from 0), forcing
                    141: speed/parity.  Because the NetBSD command line lacks a
                    142: "console=pc" argument, it will use the default "xencons" console device,
                    143: which directs the console I/O through Xen to the same console device Xen
                    144: itself uses (in this case, the serial port).
                    145: 
1.157     maxv      146: In an attempt to add performance, one can also add `dom0_max_vcpus=1 dom0_vcpus_pin`,
1.37      gdt       147: to force only one vcpu to be provided (since NetBSD dom0 can't use
1.158     maxv      148: more) and to pin that vcpu to a physical CPU. Xen has
1.178     gdt       149: [many boot options](http://xenbits.xenproject.org/docs/4.13-testing/misc/xen-command-line.html),
1.111     wiz       150: and other than dom0 memory and max_vcpus, they are generally not
1.93      gdt       151: necessary.
                    152: 
1.159     maxv      153: Copy the boot scripts into `/etc/rc.d`:
                    154: 
                    155: [[!template id=programlisting text="""
                    156: # cp /usr/pkg/share/examples/rc.d/xen* /etc/rc.d/
                    157: """]]
                    158: 
1.158     maxv      159: Enable `xencommons`:
1.31      gdt       160: 
1.157     maxv      161: [[!template id=filecontent name="/etc/rc.conf" text="""
1.152     maxv      162: xencommons=YES
                    163: """]]
1.31      gdt       164: 
1.158     maxv      165: Now, reboot so that you are running a DOM0 kernel under Xen, rather
                    166: than GENERIC without Xen.
1.157     maxv      167: 
1.31      gdt       168: TODO: Recommend for/against xen-watchdog.
1.27      jnemeth   169: 
1.158     maxv      170: Once the reboot is done, use `xl` to inspect Xen's boot messages,
1.157     maxv      171: available resources, and running domains.  For example:
1.34      gdt       172: 
1.153     maxv      173: [[!template id=programlisting text="""
                    174: # xl dmesg
                    175: ... xen's boot info ...
                    176: # xl info
                    177: ... available memory, etc ...
                    178: # xl list
                    179: Name              Id  Mem(MB)  CPU  State  Time(s)  Console
                    180: Domain-0           0       64    0  r----     58.1
                    181: """]]
                    182: 
                    183: Xen logs will be in /var/log/xen.
1.33      gdt       184: 
1.88      gdt       185: ### Issues with xencommons
                    186: 
1.157     maxv      187: `xencommons` starts `xenstored`, which stores data on behalf of dom0 and
1.88      gdt       188: domUs.  It does not currently work to stop and start xenstored.
                    189: Certainly all domUs should be shutdown first, following the sort order
                    190: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
                    191: and is not notified when xenstored exits, leading to not recreating
                    192: the state when the new xenstored starts.  Until there's a mechanism to
                    193: make this work, one should not expect to be able to restart xenstored
                    194: (and thus xencommons).  There is currently no reason to expect that
                    195: this will get fixed any time soon.
                    196: 
1.181   ! gdt       197: ## anita (for testing NetBSD)
1.41      gdt       198: 
1.178     gdt       199: With the setup so far, one should be able to run
1.82      gdt       200: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
                    201: root, because anita must create a domU):
                    202: 
1.153     maxv      203: [[!template id=programlisting text="""
                    204: anita --vmm=xl test file:///usr/obj/i386/
                    205: """]]
1.82      gdt       206: 
1.181   ! gdt       207: ## Xen-specific NetBSD issues
1.40      gdt       208: 
                    209: There are (at least) two additional things different about NetBSD as a
                    210: dom0 kernel compared to hardware.
                    211: 
1.111     wiz       212: One is that the module ABI is different because some of the #defines
1.109     gdt       213: change, so one must build modules for Xen.  As of netbsd-7, the build
1.157     maxv      214: system does this automatically.
1.40      gdt       215: 
                    216: The other difference is that XEN3_DOM0 does not have exactly the same
                    217: options as GENERIC.  While it is debatable whether or not this is a
                    218: bug, users should be aware of this and can simply add missing config
                    219: items if desired.
                    220: 
1.181   ! gdt       221: ## Updating NetBSD in a dom0
1.15      gdt       222: 
                    223: This is just like updating NetBSD on bare hardware, assuming the new
                    224: version supports the version of Xen you are running.  Generally, one
                    225: replaces the kernel and reboots, and then overlays userland binaries
1.157     maxv      226: and adjusts `/etc`.
1.15      gdt       227: 
                    228: Note that one must update both the non-Xen kernel typically used for
                    229: rescue purposes and the DOM0 kernel used with Xen.
                    230: 
1.181   ! gdt       231: ## Converting from grub to /boot
1.55      gdt       232: 
1.157     maxv      233: These instructions were used to convert a system from
1.55      gdt       234: grub to /boot.  The system was originally installed in February of
                    235: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
                    236: over time.  Before these commands, it was running NetBSD 6 i386, Xen
                    237: 4.1 and grub, much like the message linked earlier in the grub
                    238: section.
                    239: 
1.157     maxv      240: [[!template id=programlisting text="""
                    241: # Install MBR bootblocks on both disks.
                    242: fdisk -i /dev/rwd0d
                    243: fdisk -i /dev/rwd1d
                    244: # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
                    245: installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
                    246: installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
                    247: # Install secondary boot loader
                    248: cp -p /usr/mdec/boot /
                    249: # Create boot.cfg following earlier guidance:
                    250: menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
                    251: menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=512M
                    252: menu=GENERIC:boot
                    253: menu=GENERIC single-user:boot -s
                    254: menu=GENERIC.ok:boot netbsd.ok
                    255: menu=GENERIC.ok single-user:boot netbsd.ok -s
                    256: menu=Drop to boot prompt:prompt
                    257: default=1
                    258: timeout=30
                    259: """]]
1.22      gdt       260: 
1.181   ! gdt       261: ## Upgrading Xen versions
1.15      gdt       262: 
1.110     gdt       263: Minor version upgrades are trivial.  Just rebuild/replace the
1.157     maxv      264: xenkernel version and copy the new xen.gz to `/` (where `/boot.cfg`
1.110     gdt       265: references it), and reboot.
                    266: 
1.160     maxv      267: #Unprivileged domains (domU)
1.28      gdt       268: 
                    269: This section describes general concepts about domUs.  It does not
1.33      gdt       270: address specific domU operating systems or how to install them.  The
1.157     maxv      271: config files for domUs are typically in `/usr/pkg/etc/xen`, and are
1.60      wiki      272: typically named so that the file name, domU name and the domU's host
1.33      gdt       273: name match.
                    274: 
1.111     wiz       275: The domU is provided with CPU and memory by Xen, configured by the
1.33      gdt       276: dom0.  The domU is provided with disk and network by the dom0,
                    277: mediated by Xen, and configured in the dom0.
                    278: 
                    279: Entropy in domUs can be an issue; physical disks and network are on
                    280: the dom0.  NetBSD's /dev/random system works, but is often challenged.
                    281: 
1.181   ! gdt       282: ## Config files
1.48      gdt       283: 
1.165     gson      284: See /usr/pkg/share/examples/xen/xlexample*
1.164     gson      285: for a small number of well-commented examples, mostly for running
1.48      gdt       286: GNU/Linux.
                    287: 
1.157     maxv      288: The following is an example minimal domain configuration file. The domU
                    289: serves as a network file server.
                    290: 
                    291: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
                    292: name = "domU-id"
                    293: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
                    294: memory = 1024
                    295: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
                    296: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
                    297:          'file:/n0/xen/foo-wd1,0x1,w' ]
                    298: """]]
1.48      gdt       299: 
1.157     maxv      300: The domain will have name given in the `name` setting.  The kernel has the
1.48      gdt       301: host/domU name in it, so that on the dom0 one can update the various
1.157     maxv      302: domUs independently.  The `vif` line causes an interface to be provided,
1.48      gdt       303: with a specific mac address (do not reuse MAC addresses!), in bridge
                    304: mode.  Two disks are provided, and they are both writable; the bits
                    305: are stored in files and Xen attaches them to a vnd(4) device in the
1.111     wiz       306: dom0 on domain creation.  The system treats xbd0 as the boot device
1.48      gdt       307: without needing explicit configuration.
                    308: 
1.163     gson      309: By convention, domain config files are kept in `/usr/pkg/etc/xen`.  Note
1.157     maxv      310: that "xl create" takes the name of a config file, while other commands
                    311: take the name of a domain.
                    312: 
                    313: Examples of commands:
1.48      gdt       314: 
1.157     maxv      315: [[!template id=programlisting text="""
1.163     gson      316: xl create /usr/pkg/etc/xen/foo
                    317: xl console domU-id
                    318: xl create -c /usr/pkg/etc/xen/foo
                    319: xl shutdown domU-id
1.157     maxv      320: xl list
                    321: """]]
                    322: 
                    323: Typing `^]` will exit the console session.  Shutting down a domain is
1.48      gdt       324: equivalent to pushing the power button; a NetBSD domU will receive a
                    325: power-press event and do a clean shutdown.  Shutting down the dom0
                    326: will trigger controlled shutdowns of all configured domUs.
                    327: 
1.181   ! gdt       328: ## CPU and memory
1.33      gdt       329: 
1.169     plunky    330: A domain is provided with some number of vcpus, up to the number
1.157     maxv      331: of CPUs seen by the hypervisor. For a domU, it is controlled
1.48      gdt       332: from the config file by the "vcpus = N" directive.
                    333: 
                    334: A domain is provided with memory; this is controlled in the config
                    335: file by "memory = N" (in megabytes).  In the straightforward case, the
                    336: sum of the the memory allocated to the dom0 and all domUs must be less
1.33      gdt       337: than the available memory.
                    338: 
                    339: Xen also provides a "balloon" driver, which can be used to let domains
1.157     maxv      340: use more memory temporarily.
1.28      gdt       341: 
1.181   ! gdt       342: ## Virtual disks
1.28      gdt       343: 
1.158     maxv      344: In domU config files, the disks are defined as a sequence of 3-tuples:
                    345: 
                    346:  * The first element is "method:/path/to/disk". Common methods are
1.162     gutterid  347:    "file:" for a file-backed vnd, and "phy:" for something that is already
                    348:    a device, such as an LVM logical volume.
1.158     maxv      349: 
                    350:  * The second element is an artifact of how virtual disks are passed to
                    351:    Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
                    352:    are given a device name to associate with the disk, and values like
                    353:    "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
                    354:    as xbd0, the second as xbd1, and so on.  However, xl demands a
                    355:    second argument.  The name given is converted to a major/minor by
                    356:    calling stat(2) on the name in /dev and this is passed to the domU.
                    357:    In the general case, the dom0 and domU can be different operating
                    358:    systems, and it is an unwarranted assumption that they have consistent
                    359:    numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
                    360:    as both dom0 and domU, using values of 0x0 for the first disk and 0x1
                    361:    for the second works fine and avoids this issue.  For a GNU/Linux
                    362:    guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
                    363:    /dev/hda1.
1.48      gdt       364: 
1.158     maxv      365:  * The third element is "w" for writable disks, and "r" for read-only
                    366:    disks.
                    367: 
                    368: Example:
                    369: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
                    370: disk = [ 'file:/n0/xen/foo-wd0,0x0,w' ]
                    371: """]]
1.28      gdt       372: 
1.127     gdt       373: Note that NetBSD by default creates only vnd[0123].  If you need more
                    374: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
                    375: dom0.
                    376: 
                    377: Note that NetBSD by default creates only xbd[0123].  If you need more
                    378: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
                    379: 
1.28      gdt       380: Virtual Networking
                    381: ------------------
                    382: 
1.111     wiz       383: Xen provides virtual Ethernets, each of which connects the dom0 and a
1.46      gdt       384: domU.  For each virtual network, there is an interface "xvifN.M" in
1.158     maxv      385: the dom0, and a matching interface xennetM (NetBSD name) in domU index N.
                    386: The interfaces behave as if there is an Ethernet with two
1.111     wiz       387: adapters connected.  From this primitive, one can construct various
1.46      gdt       388: configurations.  We focus on two common and useful cases for which
                    389: there are existing scripts: bridging and NAT.
1.28      gdt       390: 
1.48      gdt       391: With bridging (in the example above), the domU perceives itself to be
                    392: on the same network as the dom0.  For server virtualization, this is
                    393: usually best.  Bridging is accomplished by creating a bridge(4) device
                    394: and adding the dom0's physical interface and the various xvifN.0
                    395: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
                    396: config file.  The bridge must be set up already in the dom0; an
                    397: example /etc/ifconfig.bridge0 is:
1.46      gdt       398: 
1.157     maxv      399: [[!template id=filecontent name="/etc/ifconfig.bridge0" text="""
                    400: create
                    401: up
                    402: !brconfig bridge0 add wm0
                    403: """]]
1.28      gdt       404: 
                    405: With NAT, the domU perceives itself to be behind a NAT running on the
                    406: dom0.  This is often appropriate when running Xen on a workstation.
1.48      gdt       407: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28      gdt       408: 
1.49      gdt       409: The MAC address specified is the one used for the interface in the new
1.53      gdt       410: domain.  The interface in dom0 will use this address XOR'd with
1.49      gdt       411: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
                    412: 
1.48      gdt       413: Starting domains automatically
                    414: ------------------------------
1.28      gdt       415: 
1.158     maxv      416: To start domains `domU-netbsd` and `domU-linux` at boot and shut them
                    417: down cleanly on dom0 shutdown, add the following in rc.conf:
1.28      gdt       418: 
1.158     maxv      419: [[!template id=filecontent name="/etc/rc.conf" text="""
                    420: xendomains="domU-netbsd domU-linux"
                    421: """]]
1.28      gdt       422: 
1.180     gdt       423: # Creating a domU
1.14      gdt       424: 
                    425: Creating domUs is almost entirely independent of operating system.  We
1.49      gdt       426: have already presented the basics of config files.  Note that you must
1.158     maxv      427: have already completed the dom0 setup so that "xl list" works.
1.14      gdt       428: 
1.180     gdt       429: Creating a NetBSD PV domU
                    430: --------------------------
1.1       mspo      431: 
1.49      gdt       432: See the earlier config file, and adjust memory.  Decide on how much
1.162     gutterid  433: storage you will provide, and prepare it (file or LVM).
1.49      gdt       434: 
1.111     wiz       435: While the kernel will be obtained from the dom0 file system, the same
1.49      gdt       436: file should be present in the domU as /netbsd so that tools like
                    437: savecore(8) can work.   (This is helpful but not necessary.)
                    438: 
                    439: The kernel must be specifically for Xen and for use as a domU.  The
                    440: i386 and amd64 provide the following kernels:
                    441: 
                    442:         i386 XEN3PAE_DOMU
1.95      gdt       443:         amd64 XEN3_DOMU
1.5       mspo      444: 
1.49      gdt       445: This will boot NetBSD, but this is not that useful if the disk is
                    446: empty.  One approach is to unpack sets onto the disk outside of xen
                    447: (by mounting it, just as you would prepare a physical disk for a
                    448: system you can't run the installer on).
                    449: 
                    450: A second approach is to run an INSTALL kernel, which has a miniroot
                    451: and can load sets from the network.  To do this, copy the INSTALL
                    452: kernel to / and change the kernel line in the config file to:
1.5       mspo      453: 
1.49      gdt       454:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5       mspo      455: 
1.163     gson      456: Then, start the domain as "xl create -c configfile".
1.1       mspo      457: 
1.49      gdt       458: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
                    459: line should be used in the config file.
1.1       mspo      460: 
1.3       mspo      461:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1       mspo      462: 
                    463: After booting the domain, the option to install via CDROM may be
1.49      gdt       464: selected.  The CDROM device should be changed to `xbd1d`.
1.1       mspo      465: 
1.49      gdt       466: Once done installing, "halt -p" the new domain (don't reboot or halt,
                    467: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
                    468: config file), switch the config file back to the XEN3_DOMU kernel,
                    469: and start the new domain again. Now it should be able to use "root on
                    470: xbd0a" and you should have a, functional NetBSD domU.
1.1       mspo      471: 
1.49      gdt       472: TODO: check if this is still accurate.
1.1       mspo      473: When the new domain is booting you'll see some warnings about *wscons*
                    474: and the pseudo-terminals. These can be fixed by editing the files
1.5       mspo      475: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
                    476: `/etc/ttys`, except *console*, like this:
1.1       mspo      477: 
1.3       mspo      478:     console "/usr/libexec/getty Pc"         vt100   on secure
                    479:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
                    480:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
                    481:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
                    482:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
1.1       mspo      483: 
1.5       mspo      484: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1       mspo      485: 
                    486: It is also desirable to add
                    487: 
1.49      gdt       488:         powerd=YES
1.1       mspo      489: 
1.5       mspo      490: in rc.conf. This way, the domain will be properly shut down if
1.53      gdt       491: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
1.1       mspo      492: 
1.92      gdt       493: It is not strictly necessary to have a kernel (as /netbsd) in the domU
1.111     wiz       494: file system.  However, various programs (e.g. netstat) will use that
1.92      gdt       495: kernel to look up symbols to read from kernel virtual memory.  If
                    496: /netbsd is not the running kernel, those lookups will fail.  (This is
                    497: not really a Xen-specific issue, but because the domU kernel is
                    498: obtained from the dom0, it is far more likely to be out of sync or
                    499: missing with Xen.)
                    500: 
1.158     maxv      501: Creating a Linux domU
                    502: ---------------------
1.1       mspo      503: 
                    504: Creating unprivileged Linux domains isn't much different from
                    505: unprivileged NetBSD domains, but there are some details to know.
                    506: 
                    507: First, the second parameter passed to the disk declaration (the '0x1' in
                    508: the example below)
                    509: 
1.3       mspo      510:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1       mspo      511: 
                    512: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49      gdt       513: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
                    514: So, hda1 which has major 3 and minor 1 on a Linux system will have
                    515: device number 0x301.  Alternatively, devices names can be used (hda,
                    516: hdb, ...)  as xentools has a table to map these names to devices
                    517: numbers.  To export a partition to a Linux guest we can use:
1.1       mspo      518: 
1.49      gdt       519:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
                    520:         root = "/dev/hda1 ro"
1.1       mspo      521: 
                    522: and it will appear as /dev/hda on the Linux system, and be used as root
                    523: partition.
                    524: 
1.49      gdt       525: To install the Linux system on the partition to be exported to the
                    526: guest domain, the following method can be used: install
                    527: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
                    528: that will be the root partition of your Linux domain, and mount it.
                    529: Then copy the files from a working Linux system, make adjustments in
                    530: `/etc` (fstab, network config).  It should also be possible to extract
                    531: binary packages such as .rpm or .deb directly to the mounted partition
                    532: using the appropriate tool, possibly running under NetBSD's Linux
1.111     wiz       533: emulation.  Once the file system has been populated, umount it.  If
                    534: desirable, the file system can be converted to ext3 using tune2fs -j.
1.49      gdt       535: It should now be possible to boot the Linux guest domain, using one of
                    536: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1       mspo      537: 
1.111     wiz       538: To get the Linux console right, you need to add:
1.1       mspo      539: 
1.3       mspo      540:     extra = "xencons=tty1"
1.1       mspo      541: 
1.111     wiz       542: to your configuration since not all Linux distributions auto-attach a
1.1       mspo      543: tty to the xen console.
                    544: 
1.180     gdt       545: ## Creating a NetBSD HVM domU
                    546: 
                    547: Use type='hmv', probably.  Use a GENERIC kernel within the disk image.
                    548: 
                    549: ## Creating a NetBSD PVH domU
                    550: 
                    551: Use type='pvh'.
                    552: 
                    553: \todo Explain where the kernel comes from.
                    554: 
                    555: 
1.158     maxv      556: Creating a Solaris domU
                    557: -----------------------
1.1       mspo      558: 
1.50      gdt       559: See possibly outdated
                    560: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5       mspo      561: 
1.1       mspo      562: 
1.52      gdt       563: PCI passthrough: Using PCI devices in guest domains
                    564: ---------------------------------------------------
1.1       mspo      565: 
1.180     gdt       566: NB: PCI passthrough only works on some Xen versions and as of 2020 it
                    567: is not clear that it works on any version in pkgsrc.  Reports
                    568: confirming or denying this notion should be sent to port-xen@.
                    569: 
1.53      gdt       570: The dom0 can give other domains access to selected PCI
1.52      gdt       571: devices. This can allow, for example, a non-privileged domain to have
                    572: access to a physical network interface or disk controller.  However,
                    573: keep in mind that giving a domain access to a PCI device most likely
                    574: will give the domain read/write access to the whole physical memory,
                    575: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53      gdt       576: device.  Also, it's not possible to export ISA devices to non-dom0
1.52      gdt       577: domains, which means that the primary VGA adapter can't be exported.
                    578: A guest domain trying to access the VGA registers will panic.
                    579: 
1.53      gdt       580: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52      gdt       581: not been ported to later versions at this time.
                    582: 
                    583: For a PCI device to be exported to a domU, is has to be attached to
                    584: the "pciback" driver in dom0.  Devices passed to the dom0 via the
                    585: pciback.hide boot parameter will attach to "pciback" instead of the
                    586: usual driver.  The list of devices is specified as "(bus:dev.func)",
1.5       mspo      587: where bus and dev are 2-digit hexadecimal numbers, and func a
                    588: single-digit number:
1.1       mspo      589: 
1.52      gdt       590:         pciback.hide=(00:0a.0)(00:06.0)
1.1       mspo      591: 
1.52      gdt       592: pciback devices should show up in the dom0's boot messages, and the
1.5       mspo      593: devices should be listed in the `/kern/xen/pci` directory.
1.1       mspo      594: 
1.52      gdt       595: PCI devices to be exported to a domU are listed in the "pci" array of
                    596: the domU's config file, with the format "0000:bus:dev.func".
1.1       mspo      597: 
1.52      gdt       598:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1       mspo      599: 
1.52      gdt       600: In the domU an "xpci" device will show up, to which one or more pci
1.111     wiz       601: buses will attach.  Then the PCI drivers will attach to PCI buses as
1.52      gdt       602: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
                    603: or any PCI drivers built in by default; you have to build your own
                    604: kernel to use PCI devices in a domU.  Here's a kernel config example;
                    605: note that only the "xpci" lines are unusual.
                    606: 
                    607:         include         "arch/i386/conf/XEN3_DOMU"
                    608: 
1.111     wiz       609:         # Add support for PCI buses to the XEN3_DOMU kernel
1.52      gdt       610:         xpci* at xenbus ?
                    611:         pci* at xpci ?
                    612: 
                    613:         # PCI USB controllers
                    614:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
                    615: 
                    616:         # USB bus support
                    617:         usb*    at uhci?
                    618: 
                    619:         # USB Hubs
                    620:         uhub*   at usb?
                    621:         uhub*   at uhub? port ? configuration ? interface ?
                    622: 
                    623:         # USB Mass Storage
                    624:         umass*  at uhub? port ? configuration ? interface ?
                    625:         wd*     at umass?
                    626:         # SCSI controllers
                    627:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
                    628: 
                    629:         # SCSI bus support (for both ahc and umass)
                    630:         scsibus* at scsi?
                    631: 
                    632:         # SCSI devices
                    633:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
                    634:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
1.1       mspo      635: 
                    636: 
1.173     gdt       637: # Specific Issues
                    638: 
                    639: ## domU
                    640: 
1.176     gdt       641: [NetBSD 5 is known to panic.](http://mail-index.netbsd.org/port-xen/2018/04/17/msg009181.html)
1.175     gdt       642: (However, NetBSD 5 systems should be updated to a supported version.)
1.173     gdt       643: 
                    644: # NetBSD as a domU in a VPS
1.28      gdt       645: 
                    646: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
                    647: hardware.  This section explains how to deal with Xen in a domU as a
                    648: virtual private server where you do not control or have access to the
1.70      gdt       649: dom0.  This is not intended to be an exhaustive list of VPS providers;
                    650: only a few are mentioned that specifically support NetBSD.
1.28      gdt       651: 
1.52      gdt       652: VPS operators provide varying degrees of access and mechanisms for
                    653: configuration.  The big issue is usually how one controls which kernel
1.111     wiz       654: is booted, because the kernel is nominally in the dom0 file system (to
                    655: which VPS users do not normally have access).  A second issue is how
1.70      gdt       656: to install NetBSD.
1.52      gdt       657: A VPS user may want to compile a kernel for security updates, to run
                    658: npf, run IPsec, or any other reason why someone would want to change
                    659: their kernel.
                    660: 
1.111     wiz       661: One approach is to have an administrative interface to upload a kernel,
1.68      gdt       662: or to select from a prepopulated list.  Other approaches are pygrub
1.59      gdt       663: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
1.111     wiz       664: kernel from the domU file system.  This is closer to a regular physical
1.59      gdt       665: computer, where someone who controls a machine can replace the kernel.
1.52      gdt       666: 
1.74      gdt       667: A second issue is multiple CPUs.  With NetBSD 6, domUs support
                    668: multiple vcpus, and it is typical for VPS providers to enable multiple
                    669: CPUs for NetBSD domUs.
                    670: 
1.180     gdt       671: ## Complexities due to Xen changes
                    672: 
                    673: Xen has many security advisories and people running Xen systems make
                    674: different choices.
                    675: 
                    676: ### stub domains
                    677: 
                    678: Some (Linux only?) dom0 systems use something called "stub domains" to
                    679: isolate qemu from the dom0 system, as a security and reliabilty
                    680: mechanism when running HVM domUs.  Somehow, NetBSD's GENERIC kernel
                    681: ends up using PIO for disks rather than DMA.  Of course, all of this
                    682: is emulated, but emulated PIO is unusably slow.  This problem is not
                    683: currently understood.
                    684: 
                    685: ### Grant tables
                    686: 
                    687: There are multiple versions of using grant tables, and some security
                    688: advisories have suggested disabling some versions.  Some versions of
                    689: NetBSD apparently only use specific versions and this can lead to
                    690: "NetBSD current doesn't run on hosting provider X" situations.
                    691: 
                    692: \todo Explain better.
                    693: 
1.181   ! gdt       694: ## Boot methods
        !           695: 
        !           696: ### pvgrub
1.59      gdt       697: 
                    698: pvgrub is a version of grub that uses PV operations instead of BIOS
                    699: calls.  It is booted from the dom0 as the domU kernel, and then reads
1.111     wiz       700: /grub/menu.lst and loads a kernel from the domU file system.
1.59      gdt       701: 
1.70      gdt       702: [Panix](http://www.panix.com/) lets users use pvgrub.  Panix reports
1.71      gdt       703: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
                    704: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
1.70      gdt       705: page](http://www.panix.com/v-colo/grub.html), which describes only
1.74      gdt       706: Linux but should be updated to cover NetBSD :-).
1.70      gdt       707: 
                    708: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
                    709: their own kernel.  See then [prgmr.com NetBSD
1.74      gdt       710: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
                    711: (which is in need of updating).
1.59      gdt       712: 
1.70      gdt       713: It appears that [grub's FFS
                    714: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
                    715: does not support all aspects of modern FFS, but there are also reports
1.72      gdt       716: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
1.70      gdt       717: partition for the kernel with the intent that grub can understand it,
                    718: which leads to /netbsd not being the actual kernel.  One must remember
1.111     wiz       719: to update the special boot partition.
1.59      gdt       720: 
1.181   ! gdt       721: ### pygrub
1.168     wiki      722: 
                    723: pygrub runs in the dom0 and looks into the domU file system.  This
                    724: implies that the domU must have a kernel in a file system in a format
                    725: known to pygrub.
                    726: 
                    727: pygrub doesn't seem to work to load Linux images under NetBSD dom0,
                    728: and is inherently less secure than pvgrub due to running inside dom0. For both these
                    729: reasons, pygrub should not be used, and is only still present so that
                    730: historical DomU images using it still work.
                    731: 
                    732: As of 2014, pygrub seems to be of mostly historical
                    733: interest. New DomUs should use pvgrub.
                    734: 
1.181   ! gdt       735: ## Specific Providers
        !           736: 
        !           737: ### Amazon
1.59      gdt       738: 
1.143     wiki      739: See the [Amazon EC2 page](/amazon_ec2/).

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