7: Xen is a hypervisor (or virtual machine monitor) for x86 hardware
8: (i686-class or higher), which supports running multiple guest
9: operating systems on a single physical machine. Xen is a Type 1 or
10: bare-metal hypervisor; one uses the Xen kernel to control the CPU,
11: memory and console, a dom0 operating system which mediates access to
12: other hardware (e.g., disks, network, USB), and one or more domU
13: operating systems which operate in an unprivileged virtualized
14: environment. IO requests from the domU systems are forwarded by the
15: hypervisor (Xen) to the dom0 to be fulfilled.
17: Xen supports two styles of guests. The original is Para-Virtualized
18: (PV) which means that the guest OS does not attempt to access hardware
19: directly, but instead makes hypercalls to the hypervisor. This is
20: analogous to a user-space program making system calls. (The dom0
21: operating system uses PV calls for some functions, such as updating
22: memory mapping page tables, but has direct hardware access for disk
23: and network.) PV guests must be specifically coded for Xen.
25: The more recent style is HVM, which means that the guest does not have
26: code for Xen and need not be aware that it is running under Xen.
27: Attempts to access hardware registers are trapped and emulated. This
28: style is less efficient but can run unmodified guests.
30: Generally any amd64 machine will work with Xen and PV guests. In
31: theory i386 computers without amd64 support can be used for Xen <=
32: 4.2, but we have no recent reports of this working (this is a hint).
33: For HVM guests, the VT or VMX cpu feature (Intel) or SVM/HVM/VT
34: (amd64) is needed; "cpuctl identify 0" will show this. TODO: Clean up
35: and check the above features.
37: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
38: The dom0 can start one or more domUs. (Booting is explained in detail
39: in the dom0 section.)
41: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
42: and that Xen kernels and tools are available in pkgsrc. This HOWTO
43: attempts to address both the case of running a NetBSD dom0 on hardware
44: and running domUs under it (NetBSD and other), and also running NetBSD
45: as a domU in a VPS.
47: Some versions of Xen support "PCI passthrough", which means that
48: specific PCI devices can be made available to a specific domU instead
49: of the dom0. This can be useful to let a domU run X11, or access some
50: network interface or other peripheral.
52: NetBSD used to support Xen2; this has been removed.
57: Installing NetBSD/Xen is not extremely difficult, but it is more
58: complex than a normal installation of NetBSD.
59: In general, this HOWTO is occasionally overly restrictive about how
60: things must be done, guiding the reader to stay on the established
61: path when there are no known good reasons to stray.
63: This HOWTO presumes a basic familiarity with the Xen system
64: architecture. This HOWTO presumes familiarity with installing NetBSD
65: on i386/amd64 hardware and installing software from pkgsrc.
66: See also the [Xen website](http://www.xenproject.org/).
68: Versions of Xen and NetBSD
71: Most of the installation concepts and instructions are independent
72: of Xen version and NetBSD version. This section gives advice on
73: which version to choose. Versions not in pkgsrc and older unsupported
74: versions of NetBSD are intentionally ignored.
79: In NetBSD, xen is provided in pkgsrc, via matching pairs of packages
80: xenkernel and xentools. We will refer only to the kernel versions,
81: but note that both packages must be installed together and must have
82: matching versions.
84: xenkernel3 provides Xen 3.1. This no longer receives security patches
85: and should not be used. It supports PCI passthrough, which is why
86: people use it anyway. Xen 3.1 supports non-PAE on i386.
88: xenkernel33 provides Xen 3.3. This no longer receives security
89: patches and should not be used. Xen 3.3 supports non-PAE on i386.
91: xenkernel41 provides Xen 4.1. This is no longer maintained by Xen,
92: but as of 2014-12 receives backported security patches. There are no
93: good reasons to run this version.
95: xenkernel42 provides Xen 4.2. This is no longer maintained by Xen, but
96: as of 2014-12 receives backported security patches. The only reason
97: to run this is if you need to use xm instead of xl.
99: xenkernel45 provides Xen 4.5. This is new to pkgsrc as of 2015-01 and
100: recommended for use as a conservative choice.
102: xenkernel46 provides Xen 4.6. TODO: Probably this is the recommended
105: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
107: Ideally newer versions of Xen will be added to pkgsrc.
109: Note that NetBSD support is called XEN3. It works with Xen 3 and Xen
110: 4 because the hypercall interface has been stable.
112: Xen command program
115: Early Xen used a program called xm to manipulate the system from the
116: dom0. Starting in 4.1, a replacement program with similar behavior
117: called xl is provided, but it does not work well in 4.1. In 4.2, both
118: xm and xl work fine. 4.4 is the last version that has xm. You must
119: choose one or the other, because it affects which daemons you run.
124: The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
125: reasonable choices, with more or less the same considerations for
126: non-Xen use. Therefore, netbsd-6 is recommended as the stable version
127: of the most recent release for production use. For those wanting to
128: learn Xen or without production stability concerns, netbsd-7 is likely
129: most appropriate.
131: As of NetBSD 6, a NetBSD domU will support multiple vcpus. There is
132: no SMP support for NetBSD as dom0. (The dom0 itself doesn't really
133: need SMP; the lack of support is really a problem when using a dom0 as
134: a normal computer.)
139: Xen itself can run on i386 or amd64 machines. (Practically, almost
140: any computer where one would want to run Xen today supports amd64.)
142: Xen, the dom0 kernel, and each domU kernel can be either i386 or
143: amd64. When building a xenkernel package, one obtains i386 on an i386
144: host, and amd64 on an amd64 host. If the xen kernel is i386, then the
145: dom0 kernel and all domU kernels must be i386. With an amd64 xen
146: kernel, an amd64 dom0 kernel is known to work, and an i386 dom0 kernel
147: should in theory work. An amd64 xen/dom0 is known to support both
148: i386 and amd64 domUs.
150: i386 dom0 and domU kernels must be PAE (except for Xen 3.1); these are
151: built by default. (Note that emacs (at least) fails if run on i386
152: with PAE when built without, and vice versa, presumably due to bugs in
153: the undump code.)
155: Because of the above, the standard approach is to use amd64 for the
158: Xen 4.2 is the last version to support i386 as a host. TODO: Clarify
159: if this is about the CPU, the xen kernel, or the dom0 kernel having to
160: be amd64.
166: Mostly, NetBSD as a dom0 or domU is quite stable.
167: However, there are some open PRs indicating problems.
169: - [PR 48125](http://gnats.netbsd.org/48125)
170: - [PR 47720](http://gnats.netbsd.org/47720)
172: Note also that there are issues with sparse vnd(4) instances, but
173: these are not about Xen.
178: Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
179: xl, the NetBSD 6 stable branch, and to use an amd64 kernel as the
180: dom0. Either the i386 or amd64 of NetBSD may be used as domUs.
182: Build problems
185: Ideally, all versions of Xen in pkgsrc would build on all versions of
186: NetBSD on both i386 and amd64. However, that isn't the case. Besides
187: aging code and aging compilers, qemu (included in xentools for HVM
188: support) is difficult to build. The following are known to work or FAIL:
190: xenkernel3 netbsd-5 amd64
191: xentools3 netbsd-5 amd64
192: xentools3=hvm netbsd-5 amd64 ????
193: xenkernel33 netbsd-5 amd64
194: xentools33 netbsd-5 amd64
195: xenkernel41 netbsd-5 amd64
196: xentools41 netbsd-5 amd64
197: xenkernel42 netbsd-5 amd64
198: xentools42 netbsd-5 amd64
200: xenkernel3 netbsd-6 i386 FAIL
201: xentools3 netbsd-6 i386
202: xentools3-hvm netbsd-6 i386 FAIL (dependencies fail)
203: xenkernel33 netbsd-6 i386
204: xentools33 netbsd-6 i386
205: xenkernel41 netbsd-6 i386
206: xentools41 netbsd-6 i386
207: xenkernel42 netbsd-6 i386
208: xentools42 netbsd-6 i386 *MIXED
210: (all 3 and 33 seem to FAIL)
211: xenkernel41 netbsd-7 i386
212: xentools41 netbsd-7 i386
213: xenkernel42 netbsd-7 i386
214: xentools42 netbsd-7 i386 ??FAIL
216: (*On netbsd-6 i386, there is a xentools42 in the 2014Q3 official builds,
217: but it does not build for gdt.)
219: NetBSD as a dom0
222: NetBSD can be used as a dom0 and works very well. The following
223: sections address installation, updating NetBSD, and updating Xen.
224: Note that it doesn't make sense to talk about installing a dom0 OS
225: without also installing Xen itself. We first address installing
226: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
227: NetBSD install to a dom0 install by just changing the kernel and boot
230: For experimenting with Xen, a machine with as little as 1G of RAM and
231: 100G of disk can work. For running many domUs in productions, far
232: more will be needed.
234: Styles of dom0 operation
237: There are two basic ways to use Xen. The traditional method is for
238: the dom0 to do absolutely nothing other than providing support to some
239: number of domUs. Such a system was probably installed for the sole
240: purpose of hosting domUs, and sits in a server room on a UPS.
242: The other way is to put Xen under a normal-usage computer, so that the
243: dom0 is what the computer would have been without Xen, perhaps a
244: desktop or laptop. Then, one can run domUs at will. Purists will
245: deride this as less secure than the previous approach, and for a
246: computer whose purpose is to run domUs, they are right. But Xen and a
247: dom0 (without domUs) is not meaningfully less secure than the same
248: things running without Xen. One can boot Xen or boot regular NetBSD
249: alternately with little problems, simply refraining from starting the
250: Xen daemons when not running Xen.
252: Note that NetBSD as dom0 does not support multiple CPUs. This will
253: limit the performance of the Xen/dom0 workstation approach. In theory
254: the only issue is that the "backend drivers" are not yet MPSAFE:
257: Installation of NetBSD
261: [install NetBSD/amd64](/guide/inst/)
262: just as you would if you were not using Xen.
263: However, the partitioning approach is very important.
265: If you want to use RAIDframe for the dom0, there are no special issues
266: for Xen. Typically one provides RAID storage for the dom0, and the
267: domU systems are unaware of RAID. The 2nd-stage loader bootxx_* skips
268: over a RAID1 header to find /boot from a filesystem within a RAID
269: partition; this is no different when booting Xen.
271: There are 4 styles of providing backing storage for the virtual disks
272: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
274: With raw partitions, one has a disklabel (or gpt) partition sized for
275: each virtual disk to be used by the domU. (If you are able to predict
276: how domU usage will evolve, please add an explanation to the HOWTO.
277: Seriously, needs tend to change over time.)
279: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
280: for domU disks. This is almost as efficient as raw disk partitions
281: and more flexible. Hence raw disk partitions should typically not
282: be used.
284: One can use files in the dom0 filesystem, typically created by dd'ing
285: /dev/zero to create a specific size. This is somewhat less efficient,
286: but very convenient, as one can cp the files for backup, or move them
287: between dom0 hosts.
289: Finally, in theory one can place the files backing the domU disks in a
290: SAN. (This is an invitation for someone who has done this to add a
291: HOWTO page.)
293: Installation of Xen
296: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
297: pkgsrc (or another matching pair).
298: See [the pkgsrc
299: documentation](http://www.NetBSD.org/docs/pkgsrc/) for help with pkgsrc.
301: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm. More
302: recent versions have HVM support integrated in the main xentools
303: package. It is entirely reasonable to run only PV guests.
305: Next you need to install the selected Xen kernel itself, which is
306: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz". Copy it to /.
307: For debugging, one may copy xen-debug.gz; this is conceptually similar
308: to DIAGNOSTIC and DEBUG in NetBSD. xen-debug.gz is basically only
309: useful with a serial console. Then, place a NetBSD XEN3_DOM0 kernel
310: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
311: of a NetBSD build. If using i386, use
312: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz. (If using Xen
313: 3.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen. But you
314: should not use Xen 3.1.) Both xen and the NetBSD kernel may be (and
315: typically are) left compressed.
317: In a dom0 kernel, kernfs is mandatory for xend to comunicate with the
318: kernel, so ensure that /kern is in fstab. TODO: Say this is default,
319: or file a PR and give a reference.
321: Because you already installed NetBSD, you have a working boot setup
322: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
323: beginning of your root filesystem, /boot present, and likely
324: /boot.cfg. (If not, fix before continuing!)
326: Add a line to to /boot.cfg to boot Xen. See boot.cfg(5) for an
327: example. The basic line is
329: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
331: which specifies that the dom0 should have 256M, leaving the rest to be
332: allocated for domUs. To use a serial console, use
334: menu=Xen:load /netbsd-XEN3_DOM0.gz console=com0;multiboot /xen.gz dom0_mem=256M console=com1 com1=9600,8n1
336: which will use the first serial port for Xen (which counts starting
337: from 1), forcing speed/parity, and also for NetBSD (which counts
338: starting at 0). In an attempt to add performance, one can also add
340: dom0_max_vcpus=1 dom0_vcpus_pin
342: to force only one vcpu to be provided (since NetBSD dom0 can't use
343: more) and to pin that vcpu to a physical cpu. TODO: benchmark this.
345: Xen has [many boot
347: and other tham dom0 memory and max_vcpus, they are generally not
350: As with non-Xen systems, you should have a line to boot /netbsd (a
351: kernel that works without Xen) and fallback versions of the non-Xen
352: kernel, Xen, and the dom0 kernel.
354: Now, reboot so that you are running a DOM0 kernel under Xen, rather
355: than GENERIC without Xen.
357: Using grub (historic)
360: Before NetBSD's native bootloader could support Xen, the use of
361: grub was recommended. If necessary, see the
362: [old grub information](/ports/xen/howto-grub/).
364: The [HowTo on Installing into
366: explains how to set up booting a dom0 with Xen using grub with
367: NetBSD's RAIDframe. (This is obsolete with the use of NetBSD's native
370: Configuring Xen
373: Xen logs will be in /var/log/xen.
375: Now, you have a system that will boot Xen and the dom0 kernel, but not
376: do anything else special. Make sure that you have rebooted into Xen.
377: There will be no domUs, and none can be started because you still have
378: to configure the dom0 daemons.
380: The daemons which should be run vary with Xen version and with whether
381: one is using xm or xl. The Xen 3.1 and 3.3 packages use xm. Xen 4.1
382: and higher packages use xl. While is is possible to use xm with some
383: 4.x versions (TODO: 4.1 and 4.2?), the pkgsrc-provided rc.d scripts do
384: not support this as of 2014-12-26, and thus the HOWTO does not support
385: it either. (Make sure your packages are reasonably recent.)
387: For "xm" (3.1 and 3.3), you should enable xend and xenbackendd (but
388: note that you should be using 4.x):
393: For "xl" (4.x), you should enabled xend and xencommons (xenstored).
394: Trying to boot 4.x without xencommons=YES will result in a hang; it is
395: necessary to hig ^C on the console to let the machine finish booting.
396: TODO: explain why xend is installed by the package.
400: The installation of NetBSD should already have created devices for xen
401: (xencons, xenevt), but if they are not present, create them:
403: cd /dev && sh MAKEDEV xen
405: TODO: Recommend for/against xen-watchdog.
407: After you have configured the daemons and either started them (in the
408: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
409: available resources, and running domains. An example with xl follows:
411: # xl dmesg
412: [xen's boot info]
413: # xl info
414: [available memory, etc.]
415: # xl list
416: Name Id Mem(MB) CPU State Time(s) Console
417: Domain-0 0 64 0 r---- 58.1
419: ### Issues with xencommons
421: xencommons starts xenstored, which stores data on behalf of dom0 and
422: domUs. It does not currently work to stop and start xenstored.
423: Certainly all domUs should be shutdown first, following the sort order
424: of the rc.d scripts. However, the dom0 sets up state with xenstored,
425: and is not notified when xenstored exits, leading to not recreating
426: the state when the new xenstored starts. Until there's a mechanism to
427: make this work, one should not expect to be able to restart xenstored
428: (and thus xencommons). There is currently no reason to expect that
429: this will get fixed any time soon.
431: anita (for testing NetBSD)
434: With the setup so far (assuming 4.2/xl), one should be able to run
435: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
436: root, because anita must create a domU):
438: anita --vmm=xl test file:///usr/obj/i386/
440: Alternatively, one can use --vmm=xm to use xm-based domU creation
441: instead (and must, on Xen <= 4.1). TODO: confirm that anita xl really works.
443: Xen-specific NetBSD issues
446: There are (at least) two additional things different about NetBSD as a
447: dom0 kernel compared to hardware.
449: One is that modules are not usable in DOM0 kernels, so one must
450: compile in what's needed. It's not really that modules cannot work,
451: but that modules must be built for XEN3_DOM0 because some of the
452: defines change and the normal module builds don't do this. Basically,
453: enabling Xen changes the kernel ABI, and the module build system
454: doesn't cope with this.
456: The other difference is that XEN3_DOM0 does not have exactly the same
457: options as GENERIC. While it is debatable whether or not this is a
458: bug, users should be aware of this and can simply add missing config
459: items if desired.
461: Updating NetBSD in a dom0
464: This is just like updating NetBSD on bare hardware, assuming the new
465: version supports the version of Xen you are running. Generally, one
466: replaces the kernel and reboots, and then overlays userland binaries
467: and adjusts /etc.
469: Note that one must update both the non-Xen kernel typically used for
470: rescue purposes and the DOM0 kernel used with Xen.
472: Converting from grub to /boot
475: These instructions were [TODO: will be] used to convert a system from
476: grub to /boot. The system was originally installed in February of
477: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
478: over time. Before these commands, it was running NetBSD 6 i386, Xen
479: 4.1 and grub, much like the message linked earlier in the grub
482: # Install mbr bootblocks on both disks.
483: fdisk -i /dev/rwd0d
484: fdisk -i /dev/rwd1d
485: # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
486: installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
487: installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
488: # Install secondary boot loader
489: cp -p /usr/mdec/boot /
490: # Create boog.cfg following earlier guidance:
491: menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
492: menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
494: menu=GENERIC single-user:boot -s
495: menu=GENERIC.ok:boot netbsd.ok
496: menu=GENERIC.ok single-user:boot netbsd.ok -s
497: menu=Drop to boot prompt:prompt
501: TODO: actually do this and fix it if necessary.
503: Upgrading Xen versions
506: Updating Xen is conceptually not difficult, but can run into all the
507: issues found when installing Xen. Assuming migration from 4.1 to 4.2,
508: remove the xenkernel41 and xentools41 packages and install the
509: xenkernel42 and xentools42 packages. Copy the 4.2 xen.gz to /.
511: Ensure that the contents of /etc/rc.d/xen* are correct. Specifically,
512: they must match the package you just installed and not be left over
513: from some previous installation.
515: Enable the correct set of daemons; see the configuring section above.
516: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
518: Ensure that the domU config files are valid for the new version.
519: Specifically: remove autorestart=True, and ensure that disks are
520: specified with numbers as the second argument, as the examples above
521: show, and not NetBSD device names.
523: Hardware known to work
526: Arguably, this section is misplaced, and there should be a page of
527: hardware that runs NetBSD/amd64 well, with the mostly-well-founded
528: assumption that NetBSD/xen runs fine on any modern hardware that
529: NetBSD/amd64 runs well on. Until then, we give motherboard/CPU/RAM
530: triples to aid those choosing a motherboard. Note that Xen systems
531: usually do not run X, so a listing here does not imply that X works at
534: Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
535: Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
536: ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
538: Older hardware:
540: Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
542: Running Xen under qemu
545: The astute reader will note that this section is somewhat twisted.
546: However, it can be useful to run Xen under qemu either because the
547: version of NetBSD as a dom0 does not run on the hardware in use, or to
548: generate automated test cases involving Xen.
550: In 2015-01, the following combination was reported to mostly work:
552: host OS: NetBSD/amd64 6.1.4
553: qemu: 2.2.0 from pkgsrc
554: Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
555: dom0 kernel: NetBSD/amd64 6.1.5
556: Xen tools: xentools42-4.2.5 from pkgsrc
558: See [PR 47720](http://gnats.netbsd.org/47720) for a problem with dom0
561: Unprivileged domains (domU)
564: This section describes general concepts about domUs. It does not
565: address specific domU operating systems or how to install them. The
566: config files for domUs are typically in /usr/pkg/etc/xen, and are
567: typically named so that the file name, domU name and the domU's host
568: name match.
570: The domU is provided with cpu and memory by Xen, configured by the
571: dom0. The domU is provided with disk and network by the dom0,
572: mediated by Xen, and configured in the dom0.
574: Entropy in domUs can be an issue; physical disks and network are on
575: the dom0. NetBSD's /dev/random system works, but is often challenged.
577: Config files
580: There is no good order to present config files and the concepts
581: surrounding what is being configured. We first show an example config
582: file, and then in the various sections give details.
584: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
585: for a large number of well-commented examples, mostly for running
588: The following is an example minimal domain configuration file
589: "/usr/pkg/etc/xen/foo". It is (with only a name change) an actual
590: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
591: i386 domU). The domU serves as a network file server.
593: # -*- mode: python; -*-
595: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
596: memory = 1024
597: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
598: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
599: 'file:/n0/xen/foo-wd1,0x1,w' ]
601: The domain will have the same name as the file. The kernel has the
602: host/domU name in it, so that on the dom0 one can update the various
603: domUs independently. The vif line causes an interface to be provided,
604: with a specific mac address (do not reuse MAC addresses!), in bridge
605: mode. Two disks are provided, and they are both writable; the bits
606: are stored in files and Xen attaches them to a vnd(4) device in the
607: dom0 on domain creation. The system treates xbd0 as the boot device
608: without needing explicit configuration.
610: By default xm looks for domain config files in /usr/pkg/etc/xen. Note
611: that "xm create" takes the name of a config file, while other commands
612: take the name of a domain. To create the domain, connect to the
613: console, create the domain while attaching the console, shutdown the
614: domain, and see if it has finished stopping, do (or xl with Xen >=
617: xm create foo
618: xm console foo
619: xm create -c foo
620: xm shutdown foo
621: xm list
623: Typing ^] will exit the console session. Shutting down a domain is
624: equivalent to pushing the power button; a NetBSD domU will receive a
625: power-press event and do a clean shutdown. Shutting down the dom0
626: will trigger controlled shutdowns of all configured domUs.
628: domU kernels
631: On a physical computer, the BIOS reads sector 0, and a chain of boot
632: loaders finds and loads a kernel. Normally this comes from the root
633: filesystem. With Xen domUs, the process is totally different. The
634: normal path is for the domU kernel to be a file in the dom0's
635: filesystem. At the request of the dom0, Xen loads that kernel into a
636: new domU instance and starts execution. While domU kernels can be
637: anyplace, reasonable places to store domU kernels on the dom0 are in /
638: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
639: config files), or in /u0/xen (where the vdisks are).
641: Note that loading the domU kernel from the dom0 implies that boot
642: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
643: See the VPS section near the end for discussion of alternate ways to
644: obtain domU kernels.
646: CPU and memory
649: A domain is provided with some number of vcpus, less than the number
650: of cpus seen by the hypervisor. (For a dom0, this is controlled by
651: the boot argument "dom0_max_vcpus=1".) For a domU, it is controlled
652: from the config file by the "vcpus = N" directive.
654: A domain is provided with memory; this is controlled in the config
655: file by "memory = N" (in megabytes). In the straightforward case, the
656: sum of the the memory allocated to the dom0 and all domUs must be less
657: than the available memory.
659: Xen also provides a "balloon" driver, which can be used to let domains
660: use more memory temporarily. TODO: Explain better, and explain how
661: well it works with NetBSD.
663: Virtual disks
666: With the file/vnd style, typically one creates a directory,
667: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
668: domUs. Then, for each domU disk, one writes zeros to a file that then
669: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
670: for the first virtual disk for the domU called foo. Writing zeros to
671: the file serves two purposes. One is that preallocating the contents
672: improves performance. The other is that vnd on sparse files has
673: failed to work. TODO: give working/notworking NetBSD versions for
674: sparse vnd. Note that the use of file/vnd for Xen is not really
675: different than creating a file-backed virtual disk for some other
676: purpose, except that xentools handles the vnconfig commands. To
677: create an empty 4G virtual disk, simply do
679: dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
681: Do not use qemu-img-xen, because this will create sparse file. There
682: have been recent (2015) reports of sparse vnd(4) devices causing
683: lockups, but there is apparently no PR.
685: With the lvm style, one creates logical devices. They are then used
686: similarly to vnds. TODO: Add an example with lvm.
688: In domU config files, the disks are defined as a sequence of 3-tuples.
689: The first element is "method:/path/to/disk". Common methods are
690: "file:" for file-backed vnd. and "phy:" for something that is already
691: a (TODO: character or block) device.
693: The second element is an artifact of how virtual disks are passed to
694: Linux, and a source of confusion with NetBSD Xen usage. Linux domUs
695: are given a device name to associate with the disk, and values like
696: "hda1" or "sda1" are common. In a NetBSD domU, the first disk appears
697: as xbd0, the second as xbd1, and so on. However, xm/xl demand a
698: second argument. The name given is converted to a major/minor by
699: calling stat(2) on the name in /dev and this is passed to the domU.
700: In the general case, the dom0 and domU can be different operating
701: systems, and it is an unwarranted assumption that they have consistent
702: numbering in /dev, or even that the dom0 OS has a /dev. With NetBSD
703: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
704: for the second works fine and avoids this issue. For a GNU/Linux
705: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
708: The third element is "w" for writable disks, and "r" for read-only
711: Virtual Networking
714: Xen provides virtual ethernets, each of which connects the dom0 and a
715: domU. For each virtual network, there is an interface "xvifN.M" in
716: the dom0, and in domU index N, a matching interface xennetM (NetBSD
717: name). The interfaces behave as if there is an Ethernet with two
718: adaptors connected. From this primitive, one can construct various
719: configurations. We focus on two common and useful cases for which
720: there are existing scripts: bridging and NAT.
722: With bridging (in the example above), the domU perceives itself to be
723: on the same network as the dom0. For server virtualization, this is
724: usually best. Bridging is accomplished by creating a bridge(4) device
725: and adding the dom0's physical interface and the various xvifN.0
726: interfaces to the bridge. One specifies "bridge=bridge0" in the domU
727: config file. The bridge must be set up already in the dom0; an
728: example /etc/ifconfig.bridge0 is:
732: !brconfig bridge0 add wm0
734: With NAT, the domU perceives itself to be behind a NAT running on the
735: dom0. This is often appropriate when running Xen on a workstation.
736: TODO: NAT appears to be configured by "vif = [ '' ]".
738: The MAC address specified is the one used for the interface in the new
739: domain. The interface in dom0 will use this address XOR'd with
740: 00:00:00:01:00:00. Random MAC addresses are assigned if not given.
742: Sizing domains
745: Modern x86 hardware has vast amounts of resources. However, many
746: virtual servers can function just fine on far less. A system with
747: 256M of RAM and a 4G disk can be a reasonable choice. Note that it is
748: far easier to adjust virtual resources than physical ones. For
749: memory, it's just a config file edit and a reboot. For disk, one can
750: create a new file and vnconfig it (or lvm), and then dump/restore,
751: just like updating physical disks, but without having to be there and
752: without those pesky connectors.
754: Starting domains automatically
757: To start domains foo at bar at boot and shut them down cleanly on dom0
758: shutdown, in rc.conf add:
760: xendomains="foo bar"
762: Note that earlier versions of the xentools41 xendomains rc.d scripth
763: usd xl, when one should use xm with 4.1.
765: Creating specific unprivileged domains (domU)
768: Creating domUs is almost entirely independent of operating system. We
769: have already presented the basics of config files. Note that you must
770: have already completed the dom0 setup so that "xl list" (or "xm list")
773: Creating an unprivileged NetBSD domain (domU)
776: See the earlier config file, and adjust memory. Decide on how much
777: storage you will provide, and prepare it (file or lvm).
779: While the kernel will be obtained from the dom0 filesystem, the same
780: file should be present in the domU as /netbsd so that tools like
781: savecore(8) can work. (This is helpful but not necessary.)
783: The kernel must be specifically for Xen and for use as a domU. The
784: i386 and amd64 provide the following kernels:
786: i386 XEN3_DOMU
787: i386 XEN3PAE_DOMU
788: amd64 XEN3_DOMU
790: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
791: use the PAE version of the i386 kernel.
793: This will boot NetBSD, but this is not that useful if the disk is
794: empty. One approach is to unpack sets onto the disk outside of xen
795: (by mounting it, just as you would prepare a physical disk for a
796: system you can't run the installer on).
798: A second approach is to run an INSTALL kernel, which has a miniroot
799: and can load sets from the network. To do this, copy the INSTALL
800: kernel to / and change the kernel line in the config file to:
802: kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
804: Then, start the domain as "xl create -c configname".
806: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
807: line should be used in the config file.
809: disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
811: After booting the domain, the option to install via CDROM may be
812: selected. The CDROM device should be changed to `xbd1d`.
814: Once done installing, "halt -p" the new domain (don't reboot or halt,
815: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
816: config file), switch the config file back to the XEN3_DOMU kernel,
817: and start the new domain again. Now it should be able to use "root on
818: xbd0a" and you should have a, functional NetBSD domU.
820: TODO: check if this is still accurate.
821: When the new domain is booting you'll see some warnings about *wscons*
822: and the pseudo-terminals. These can be fixed by editing the files
823: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
824: `/etc/ttys`, except *console*, like this:
826: console "/usr/libexec/getty Pc" vt100 on secure
827: ttyE0 "/usr/libexec/getty Pc" vt220 off secure
828: ttyE1 "/usr/libexec/getty Pc" vt220 off secure
829: ttyE2 "/usr/libexec/getty Pc" vt220 off secure
830: ttyE3 "/usr/libexec/getty Pc" vt220 off secure
832: Finally, all screens must be commented out from `/etc/wscons.conf`.
834: It is also desirable to add
838: in rc.conf. This way, the domain will be properly shut down if
839: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
841: It is not strictly necessary to have a kernel (as /netbsd) in the domU
842: filesystem. However, various programs (e.g. netstat) will use that
843: kernel to look up symbols to read from kernel virtual memory. If
844: /netbsd is not the running kernel, those lookups will fail. (This is
845: not really a Xen-specific issue, but because the domU kernel is
846: obtained from the dom0, it is far more likely to be out of sync or
847: missing with Xen.)
849: Creating an unprivileged Linux domain (domU)
852: Creating unprivileged Linux domains isn't much different from
853: unprivileged NetBSD domains, but there are some details to know.
855: First, the second parameter passed to the disk declaration (the '0x1' in
856: the example below)
858: disk = [ 'phy:/dev/wd0e,0x1,w' ]
860: does matter to Linux. It wants a Linux device number here (e.g. 0x300
861: for hda). Linux builds device numbers as: (major \<\< 8 + minor).
862: So, hda1 which has major 3 and minor 1 on a Linux system will have
863: device number 0x301. Alternatively, devices names can be used (hda,
864: hdb, ...) as xentools has a table to map these names to devices
865: numbers. To export a partition to a Linux guest we can use:
867: disk = [ 'phy:/dev/wd0e,0x300,w' ]
868: root = "/dev/hda1 ro"
870: and it will appear as /dev/hda on the Linux system, and be used as root
873: To install the Linux system on the partition to be exported to the
874: guest domain, the following method can be used: install
875: sysutils/e2fsprogs from pkgsrc. Use mke2fs to format the partition
876: that will be the root partition of your Linux domain, and mount it.
877: Then copy the files from a working Linux system, make adjustments in
878: `/etc` (fstab, network config). It should also be possible to extract
879: binary packages such as .rpm or .deb directly to the mounted partition
880: using the appropriate tool, possibly running under NetBSD's Linux
881: emulation. Once the filesystem has been populated, umount it. If
882: desirable, the filesystem can be converted to ext3 using tune2fs -j.
883: It should now be possible to boot the Linux guest domain, using one of
884: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
886: To get the linux console right, you need to add:
888: extra = "xencons=tty1"
890: to your configuration since not all linux distributions auto-attach a
891: tty to the xen console.
893: Creating an unprivileged Solaris domain (domU)
896: See possibly outdated
897: [Solaris domU instructions](/ports/xen/howto-solaris/).
900: PCI passthrough: Using PCI devices in guest domains
903: The dom0 can give other domains access to selected PCI
904: devices. This can allow, for example, a non-privileged domain to have
905: access to a physical network interface or disk controller. However,
906: keep in mind that giving a domain access to a PCI device most likely
907: will give the domain read/write access to the whole physical memory,
908: as PCs don't have an IOMMU to restrict memory access to DMA-capable
909: device. Also, it's not possible to export ISA devices to non-dom0
910: domains, which means that the primary VGA adapter can't be exported.
911: A guest domain trying to access the VGA registers will panic.
913: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
914: not been ported to later versions at this time.
916: For a PCI device to be exported to a domU, is has to be attached to
917: the "pciback" driver in dom0. Devices passed to the dom0 via the
918: pciback.hide boot parameter will attach to "pciback" instead of the
919: usual driver. The list of devices is specified as "(bus:dev.func)",
920: where bus and dev are 2-digit hexadecimal numbers, and func a
921: single-digit number:
925: pciback devices should show up in the dom0's boot messages, and the
926: devices should be listed in the `/kern/xen/pci` directory.
928: PCI devices to be exported to a domU are listed in the "pci" array of
929: the domU's config file, with the format "0000:bus:dev.func".
931: pci = [ '0000:00:06.0', '0000:00:0a.0' ]
933: In the domU an "xpci" device will show up, to which one or more pci
934: busses will attach. Then the PCI drivers will attach to PCI busses as
935: usual. Note that the default NetBSD DOMU kernels do not have "xpci"
936: or any PCI drivers built in by default; you have to build your own
937: kernel to use PCI devices in a domU. Here's a kernel config example;
938: note that only the "xpci" lines are unusual.
940: include "arch/i386/conf/XEN3_DOMU"
942: # Add support for PCI busses to the XEN3_DOMU kernel
943: xpci* at xenbus ?
944: pci* at xpci ?
946: # PCI USB controllers
947: uhci* at pci? dev ? function ? # Universal Host Controller (Intel)
949: # USB bus support
950: usb* at uhci?
952: # USB Hubs
953: uhub* at usb?
954: uhub* at uhub? port ? configuration ? interface ?
956: # USB Mass Storage
957: umass* at uhub? port ? configuration ? interface ?
958: wd* at umass?
959: # SCSI controllers
960: ahc* at pci? dev ? function ? # Adaptec 94x, aic78x0 SCSI
962: # SCSI bus support (for both ahc and umass)
963: scsibus* at scsi?
965: # SCSI devices
966: sd* at scsibus? target ? lun ? # SCSI disk drives
967: cd* at scsibus? target ? lun ? # SCSI CD-ROM drives
970: NetBSD as a domU in a VPS
973: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
974: hardware. This section explains how to deal with Xen in a domU as a
975: virtual private server where you do not control or have access to the
976: dom0. This is not intended to be an exhaustive list of VPS providers;
977: only a few are mentioned that specifically support NetBSD.
979: VPS operators provide varying degrees of access and mechanisms for
980: configuration. The big issue is usually how one controls which kernel
981: is booted, because the kernel is nominally in the dom0 filesystem (to
982: which VPS users do not normally have acesss). A second issue is how
983: to install NetBSD.
984: A VPS user may want to compile a kernel for security updates, to run
985: npf, run IPsec, or any other reason why someone would want to change
986: their kernel.
988: One approach is to have an adminstrative interface to upload a kernel,
989: or to select from a prepopulated list. Other approaches are pygrub
990: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
991: kernel from the domU filesystem. This is closer to a regular physical
992: computer, where someone who controls a machine can replace the kernel.
994: A second issue is multiple CPUs. With NetBSD 6, domUs support
995: multiple vcpus, and it is typical for VPS providers to enable multiple
996: CPUs for NetBSD domUs.
1001: pygrub runs in the dom0 and looks into the domU filesystem. This
1002: implies that the domU must have a kernel in a filesystem in a format
1003: known to pygrub. As of 2014, pygrub seems to be of mostly historical
1009: pvgrub is a version of grub that uses PV operations instead of BIOS
1010: calls. It is booted from the dom0 as the domU kernel, and then reads
1011: /grub/menu.lst and loads a kernel from the domU filesystem.
1013: [Panix](http://www.panix.com/) lets users use pvgrub. Panix reports
1014: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
1015: (and hence with defaults from "newfs -O 2"). See [Panix's pvgrub
1016: page](http://www.panix.com/v-colo/grub.html), which describes only
1017: Linux but should be updated to cover NetBSD :-).
1019: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
1020: their own kernel. See then [prgmr.com NetBSD
1022: (which is in need of updating).
1024: It appears that [grub's FFS
1026: does not support all aspects of modern FFS, but there are also reports
1027: that FFSv2 works fine. At prgmr, typically one has an ext2 or FAT
1028: partition for the kernel with the intent that grub can understand it,
1029: which leads to /netbsd not being the actual kernel. One must remember
1030: to update the special boot partiion.
1035: See the [Amazon EC2 page](../amazon_ec2/).
1037: Using npf
1040: In standard kernels, npf is a module, and thus cannot be loaded in a
1041: DOMU kernel.
1043: TODO: Explain how to compile npf into a custom kernel, answering (but
1044: note that the problem was caused by not booting the right kernel)
1045: [this email to
1048: TODO items for improving NetBSD/xen
1051: * Make the NetBSD dom0 kernel work with SMP.
1052: * Test the Xen 4.5 packages adequately to be able to recommend them as
1053: the standard approach.
1054: * Get PCI passthrough working on Xen 4.5
1055: * Get pvgrub into pkgsrc, either via xentools or separately.
1056: * grub
1057: * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
1058: fragsize/blocksize (UFS2 support may be present; the point is to
1059: make it so that with any UFS1/UFS2 filesystem setup that works
1060: with NetBSD grub will also work).
1061: See [pkg/40258](http://gnats.netbsd.org/40258).
1062: * Push patches upstream.
1063: * Get UFS2 patches into pvgrub.
1064: * Add support for PV ops to a version of /boot, and make it usable as
1065: a kernel in Xen, similar to pvgrub.
1066: * Solve somehow the issue with modules for GENERIC not being loadable
1067: in a Xen dom0 or domU kernel.
1069: Random pointers
1072: TODO: This section contains links from elsewhere not yet integrated
1073: into the HOWTO.
1075: * http://www.lumbercartel.ca/library/xen/
1076: * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html
CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb