1: [[!meta title="Xen HowTo"]]
2:
3: Introduction
4: ============
5:
6: [![[Xen
7: screenshot]](https://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](https://www.netbsd.org/gallery/in-Action/hubertf-xen.png)
8:
9: Xen is a hypervisor for x86 hardware, which supports running multiple guest
10: operating systems on a single physical machine. Xen is a Type 1 or
11: bare-metal hypervisor; one uses the Xen kernel to control the CPU,
12: memory and console, a dom0 operating system which mediates access to
13: other hardware (e.g., disks, network, USB), and one or more domU
14: operating systems which operate in an unprivileged virtualized
15: environment. IO requests from the domU systems are forwarded by the
16: Xen hypervisor to the dom0 to be fulfilled.
17:
18: Xen supports different styles of guest:
19:
20: [[!table data="""
21: Style of guest |Supported by NetBSD
22: PV |Yes (dom0, domU)
23: HVM |Yes (domU)
24: PVHVM |No
25: PVH |No
26: """]]
27:
28: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
29: hardware directly, but instead makes hypercalls to the hypervisor; PV
30: guests must be specifically coded for Xen. In HVM mode, no guest
31: modification is required; however, hardware support is required, such
32: as VT-x on Intel CPUs and SVM on AMD CPUs.
33:
34: There are further features for IOMMU virtualization, Intel's VT-d and
35: AMD's AMD-Vi. TODO: Explain whether Xen on NetBSD makes use of these
36: features. TODO: Review by someone who really understands this.
37:
38: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
39: The dom0 can start one or more domUs. (Booting is explained in detail
40: in the dom0 section.)
41:
42: This HOWTO presumes a basic familiarity with the Xen system
43: architecture, with installing NetBSD on i386/amd64 hardware, and with
44: installing software from pkgsrc. See also the [Xen
45: website](http://www.xenproject.org/).
46:
47: This HOWTO attempts to address both the case of running a NetBSD dom0
48: on hardware and running domUs under it (NetBSD and other), and also
49: running NetBSD as a domU in a VPS.
50:
51: Versions of Xen and NetBSD
52: ==========================
53:
54: Most of the installation concepts and instructions are independent
55: of Xen version and NetBSD version. This section gives advice on
56: which version to choose. Versions not in pkgsrc and older unsupported
57: versions of NetBSD are intentionally ignored.
58:
59: The term "amd64" is used to refer to both the NetBSD port and to the
60: hardware architecture on which it runs. Such hardware is generally
61: made by both Intel and AMD, and common on PC computers.
62:
63: Xen versions
64: ------------
65:
66: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
67: xenkernel and xentools. We will refer only to the kernel versions,
68: but note that both packages must be installed together and must have
69: matching versions.
70:
71: Versions available in pkgsrc:
72:
73: [[!table data="""
74: Xen Version |Package Name |Xen CPU Support |EOL'ed By Upstream
75: 4.2 |xenkernel42 |32bit, 64bit |Yes
76: 4.5 |xenkernel45 |64bit |Yes
77: 4.6 |xenkernel46 |64bit |Partially
78: 4.8 |xenkernel48 |64bit |No
79: 4.11 |xenkernel411 |64bit |No
80: """]]
81:
82: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
83:
84: Note: Xen 4.2 was the last version to support 32bit CPUs.
85:
86: Xen command program
87: -------------------
88:
89: Early Xen used a program called xm to manipulate the system from the
90: dom0. Starting in 4.1, a replacement program with similar behavior
91: called xl is provided, but it does not work well in 4.1. In 4.2, both
92: xm and xl work fine. 4.4 is the last version that has xm.
93:
94: You must make a global choice to use xm or xl, because it affects not
95: only which command you use, but the command used by rc.d scripts
96: (specifically xendomains) and which daemons should be run. The
97: xentools packages provide xl for 4.2 and up.
98:
99: In 4.2, you can choose to use xm by simply changing the ctl_command
100: variable and setting xend=YES in rc.conf.
101:
102: With xl, virtual devices are configured in parallel, which can cause
103: problems if they are written assuming serial operation (e.g., updating
104: firewall rules without explicit locking). There is now locking for
105: the provided scripts, which works for normal casses (e.g, file-backed
106: xbd, where a vnd must be allocated). But, as of 201612, it has not
107: been adequately tested for a complex custom setup with a large number
108: of interfaces.
109:
110: NetBSD versions
111: ---------------
112:
113: The netbsd-7, netbsd-8, and -current branches are all reasonable
114: choices, with more or less the same considerations for non-Xen use.
115: NetBSD 8 is recommended as the stable version of the most recent
116: release for production use.
117:
118: For developing Xen, netbsd-current may be appropriate.
119:
120: As of NetBSD 6, a NetBSD domU will support multiple vcpus. There is
121: no SMP support for NetBSD as dom0. (The dom0 itself doesn't really
122: need SMP for dom0 functions; the lack of support is really a problem
123: when using a dom0 as a normal computer.)
124:
125: Note: NetBSD support is called XEN3. However, it does support Xen 4,
126: because the hypercall interface has remained identical.
127:
128: Architecture
129: ------------
130:
131: Xen itself can run on i386 (Xen < 4.2) or amd64 hardware (all Xen
132: versions). (Practically, almost any computer where one would want to
133: run Xen today supports amd64.)
134:
135: Xen, the dom0 system, and each domU system can be either i386 or
136: amd64. When building a xenkernel package, one obtains an i386 Xen
137: kernel on an i386 host, and an amd64 Xen kernel on an amd64 host. If
138: the Xen kernel is i386, then the dom0 kernel and all domU kernels must
139: be i386. With an amd64 Xen kernel, an amd64 dom0 kernel is known to
140: work, and an i386 dom0 kernel should in theory work. An amd64
141: Xen/dom0 is known to support both i386 and amd64 domUs.
142:
143: i386 dom0 and domU kernels must be PAE (except for an i386 Xen 3.1
144: kernel, where one can use non-PAE for dom0 and all domUs); PAE kernels
145: are included in the NetBSD default build. (Note that emacs (at least)
146: fails if run on i386 with PAE when built without, and vice versa,
147: presumably due to bugs in the undump code.)
148:
149: Because of the above, the standard approach is to use an amd64 Xen
150: kernel and NetBSD/amd64 for the dom0. For domUs, NetBSD/i386 (with
151: the PAE kernel) and NetBSD/amd64 are in widespread use, and there is
152: little to no Xen-specific reason to prefer one over the other.
153:
154: Note that to use an i386 dom0 with Xen 4.5 or higher, one must build
155: (or obtain from pre-built packages) an amd64 Xen kernel and install
156: that on the system. (One must also use a PAE i386 kernel, but this is
157: also required with an i386 Xen kernel.). Almost no one in the
158: NetBSD/Xen community does this, and the standard, well-tested,
159: approach is to use an amd64 dom0.
160:
161: A [posting on
162: xen-devel](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html)
163: explained that PV system call overhead was higher on amd64, and thus
164: there is some notion that i386 guests are faster. It goes on to
165: caution that the total situation is complex and not entirely
166: understood. On top of that caution, the post is about Linux, not
167: NetBSD. TODO: Include link to benchmarks, if someone posts them.
168:
169: NetBSD as a dom0
170: ================
171:
172: NetBSD can be used as a dom0 and works very well. The following
173: sections address installation, updating NetBSD, and updating Xen.
174: Note that it doesn't make sense to talk about installing a dom0 OS
175: without also installing Xen itself. We first address installing
176: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
177: NetBSD install to a dom0 install by just changing the kernel and boot
178: configuration.
179:
180: For experimenting with Xen, a machine with as little as 1G of RAM and
181: 100G of disk can work. For running many domUs in productions, far
182: more will be needed; e.g. 4-8G and 1T of disk is reasonable for a
183: half-dozen domUs of 512M and 32G each. Basically, the RAM and disk
184: have to be bigger than the sum of the RAM/disk needs of the dom0 and
185: all the domUs.
186:
187: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
188: 512M it worked reliably. This does not make sense, but if you see
189: "not ELF" after Xen boots, try increasing dom0 RAM.
190:
191: Styles of dom0 operation
192: ------------------------
193:
194: There are two basic ways to use Xen. The traditional method is for
195: the dom0 to do absolutely nothing other than providing support to some
196: number of domUs. Such a system was probably installed for the sole
197: purpose of hosting domUs, and sits in a server room on a UPS.
198:
199: The other way is to put Xen under a normal-usage computer, so that the
200: dom0 is what the computer would have been without Xen, perhaps a
201: desktop or laptop. Then, one can run domUs at will. Purists will
202: deride this as less secure than the previous approach, and for a
203: computer whose purpose is to run domUs, they are right. But Xen and a
204: dom0 (without domUs) is not meaningfully less secure than the same
205: things running without Xen. One can boot Xen or boot regular NetBSD
206: alternately with little problems, simply refraining from starting the
207: Xen daemons when not running Xen.
208:
209: Note that NetBSD as dom0 does not support multiple CPUs. This will
210: limit the performance of the Xen/dom0 workstation approach. In theory
211: the only issue is that the "backend drivers" are not yet MPSAFE:
212: https://mail-index.netbsd.org/netbsd-users/2014/08/29/msg015195.html
213:
214: Installation of NetBSD
215: ----------------------
216:
217: First,
218: [install NetBSD/amd64](/guide/inst/)
219: just as you would if you were not using Xen.
220: However, the partitioning approach is very important.
221:
222: If you want to use RAIDframe for the dom0, there are no special issues
223: for Xen. Typically one provides RAID storage for the dom0, and the
224: domU systems are unaware of RAID. The 2nd-stage loader bootxx_* skips
225: over a RAID1 header to find /boot from a file system within a RAID
226: partition; this is no different when booting Xen.
227:
228: There are 4 styles of providing backing storage for the virtual disks
229: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
230:
231: With raw partitions, one has a disklabel (or gpt) partition sized for
232: each virtual disk to be used by the domU. (If you are able to predict
233: how domU usage will evolve, please add an explanation to the HOWTO.
234: Seriously, needs tend to change over time.)
235:
236: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
237: for domU disks. This is almost as efficient as raw disk partitions
238: and more flexible. Hence raw disk partitions should typically not
239: be used.
240:
241: One can use files in the dom0 file system, typically created by dd'ing
242: /dev/zero to create a specific size. This is somewhat less efficient,
243: but very convenient, as one can cp the files for backup, or move them
244: between dom0 hosts.
245:
246: Finally, in theory one can place the files backing the domU disks in a
247: SAN. (This is an invitation for someone who has done this to add a
248: HOWTO page.)
249:
250: Installation of Xen
251: -------------------
252:
253: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
254: pkgsrc (or another matching pair). See [the pkgsrc
255: documentation](https://www.NetBSD.org/docs/pkgsrc/) for help with
256: pkgsrc. Ensure that your packages are recent; the HOWTO does not
257: contemplate old builds.
258:
259:
260: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm. More
261: recent versions have HVM support integrated in the main xentools
262: package. It is entirely reasonable to run only PV guests.
263:
264: Next you need to install the selected Xen kernel itself, which is
265: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz". Copy it to /.
266: For debugging, one may copy xen-debug.gz; this is conceptually similar
267: to DIAGNOSTIC and DEBUG in NetBSD. xen-debug.gz is basically only
268: useful with a serial console. Then, place a NetBSD XEN3_DOM0 kernel
269: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
270: of a NetBSD build. If using i386, use
271: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz. (If using Xen
272: 3.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen. But you
273: should not use Xen 3.1.) Both xen and the NetBSD kernel may be (and
274: typically are) left compressed.
275:
276: In a dom0, kernfs is mandatory for xend to communicate with the
277: kernel, so ensure that /kern is in fstab. (A standard NetBSD install
278: should already mount /kern.)
279:
280: Because you already installed NetBSD, you have a working boot setup
281: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
282: beginning of your root file system, have /boot, and likely also
283: /boot.cfg. (If not, fix before continuing!)
284:
285: Add a line to to /boot.cfg to boot Xen. See boot.cfg(5) for an
286: example. The basic line is
287:
288: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
289:
290: which specifies that the dom0 should have 512M, leaving the rest to be
291: allocated for domUs. To use a serial console, use
292:
293: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
294:
295: which will use the first serial port for Xen (which counts starting
296: from 1, unlike NetBSD which counts starting from 0), forcing
297: speed/parity. Because the NetBSD command line lacks a
298: "console=pc" argument, it will use the default "xencons" console device,
299: which directs the console I/O through Xen to the same console device Xen
300: itself uses (in this case, the serial port).
301:
302: In an attempt to add performance, one can also add
303:
304: dom0_max_vcpus=1 dom0_vcpus_pin
305:
306: to force only one vcpu to be provided (since NetBSD dom0 can't use
307: more) and to pin that vcpu to a physical CPU. TODO: benchmark this.
308:
309: Xen has [many boot
310: options](http://xenbits.xenproject.org/docs/4.5-testing/misc/xen-command-line.html),
311: and other than dom0 memory and max_vcpus, they are generally not
312: necessary.
313:
314: As with non-Xen systems, you should have a line to boot /netbsd (a
315: kernel that works without Xen). Consider a line to boot /netbsd.ok (a
316: fallback version of the non-Xen kernel, updated manually when you are
317: sure /netbsd is ok). Consider also a line to boot fallback versions
318: of Xen and the dom0 kernel, but note that non-Xen NetBSD can be used
319: to resolve Xen booting issues.
320:
321: Probably you want a default=N line to choose Xen in the absence of
322: intervention.
323:
324: Now, reboot so that you are running a DOM0 kernel under Xen, rather
325: than GENERIC without Xen.
326:
327: Using grub (historic)
328: ---------------------
329:
330: Before NetBSD's native bootloader could support Xen, the use of
331: grub was recommended. If necessary, see the
332: [old grub information](/ports/xen/howto-grub).
333:
334: The [HowTo on Installing into
335: RAID-1](https://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
336: explains how to set up booting a dom0 with Xen using grub with
337: NetBSD's RAIDframe. (This is obsolete with the use of NetBSD's native
338: boot. Now, just create a system with RAID-1, and alter /boot.cfg as
339: described above.)
340:
341: Configuring Xen
342: ---------------
343:
344: Xen logs will be in /var/log/xen.
345:
346: Now, you have a system that will boot Xen and the dom0 kernel, but not
347: do anything else special. Make sure that you have rebooted into Xen.
348: There will be no domUs, and none can be started because you still have
349: to configure the dom0 daemons.
350:
351: The daemons which should be run vary with Xen version and with whether
352: one is using xm or xl. The Xen 3.1, 3.3 and 4.1 packages use xm. Xen
353: 4.2 and up packages use xl. To use xm with 4.2, edit xendomains to
354: use xm instead.
355:
356: For 3.1 and 3.3, you should enable xend and xenbackendd:
357:
358: xend=YES
359: xenbackendd=YES
360:
361: For 4.1 and up, you should enable xencommons. Not enabling xencommons
362: will result in a hang; it is necessary to hit ^C on the console to let
363: the machine finish booting. If you are using xm (default in 4.1, or
364: if you changed xendomains in 4.2), you should also enable xend:
365:
366: xend=YES # only if using xm, and only installed <= 4.2
367: xencommons=YES
368:
369: TODO: Recommend for/against xen-watchdog.
370:
371: After you have configured the daemons and either started them (in the
372: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
373: available resources, and running domains. An example with xl follows:
374:
375: # xl dmesg
376: [xen's boot info]
377: # xl info
378: [available memory, etc.]
379: # xl list
380: Name Id Mem(MB) CPU State Time(s) Console
381: Domain-0 0 64 0 r---- 58.1
382:
383: ### Issues with xencommons
384:
385: xencommons starts xenstored, which stores data on behalf of dom0 and
386: domUs. It does not currently work to stop and start xenstored.
387: Certainly all domUs should be shutdown first, following the sort order
388: of the rc.d scripts. However, the dom0 sets up state with xenstored,
389: and is not notified when xenstored exits, leading to not recreating
390: the state when the new xenstored starts. Until there's a mechanism to
391: make this work, one should not expect to be able to restart xenstored
392: (and thus xencommons). There is currently no reason to expect that
393: this will get fixed any time soon.
394:
395: ### No-longer needed advice about devices
396:
397: The installation of NetBSD should already have created devices for xen
398: (xencons, xenevt, xsd_kva), but if they are not present, create them:
399:
400: cd /dev && sh MAKEDEV xen
401:
402: anita (for testing NetBSD)
403: --------------------------
404:
405: With the setup so far (assuming 4.2/xl), one should be able to run
406: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
407: root, because anita must create a domU):
408:
409: anita --vmm=xl test file:///usr/obj/i386/
410:
411: Alternatively, one can use --vmm=xm to use xm-based domU creation
412: instead (and must, on Xen <= 4.1). TODO: confirm that anita xl really works.
413:
414: Xen-specific NetBSD issues
415: --------------------------
416:
417: There are (at least) two additional things different about NetBSD as a
418: dom0 kernel compared to hardware.
419:
420: One is that the module ABI is different because some of the #defines
421: change, so one must build modules for Xen. As of netbsd-7, the build
422: system does this automatically. TODO: check this. (Before building
423: Xen modules was added, it was awkward to use modules to the point
424: where it was considered that it did not work.)
425:
426: The other difference is that XEN3_DOM0 does not have exactly the same
427: options as GENERIC. While it is debatable whether or not this is a
428: bug, users should be aware of this and can simply add missing config
429: items if desired.
430:
431: Updating NetBSD in a dom0
432: -------------------------
433:
434: This is just like updating NetBSD on bare hardware, assuming the new
435: version supports the version of Xen you are running. Generally, one
436: replaces the kernel and reboots, and then overlays userland binaries
437: and adjusts /etc.
438:
439: Note that one must update both the non-Xen kernel typically used for
440: rescue purposes and the DOM0 kernel used with Xen.
441:
442: Converting from grub to /boot
443: -----------------------------
444:
445: These instructions were [TODO: will be] used to convert a system from
446: grub to /boot. The system was originally installed in February of
447: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
448: over time. Before these commands, it was running NetBSD 6 i386, Xen
449: 4.1 and grub, much like the message linked earlier in the grub
450: section.
451:
452: # Install MBR bootblocks on both disks.
453: fdisk -i /dev/rwd0d
454: fdisk -i /dev/rwd1d
455: # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
456: installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
457: installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
458: # Install secondary boot loader
459: cp -p /usr/mdec/boot /
460: # Create boot.cfg following earlier guidance:
461: menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
462: menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=512M
463: menu=GENERIC:boot
464: menu=GENERIC single-user:boot -s
465: menu=GENERIC.ok:boot netbsd.ok
466: menu=GENERIC.ok single-user:boot netbsd.ok -s
467: menu=Drop to boot prompt:prompt
468: default=1
469: timeout=30
470:
471: TODO: actually do this and fix it if necessary.
472:
473: Upgrading Xen versions
474: ---------------------
475:
476: Minor version upgrades are trivial. Just rebuild/replace the
477: xenkernel version and copy the new xen.gz to / (where /boot.cfg
478: references it), and reboot.
479:
480: Major version upgrades are conceptually not difficult, but can run
481: into all the issues found when installing Xen. Assuming migration
482: from 4.1 to 4.2, remove the xenkernel41 and xentools41 packages and
483: install the xenkernel42 and xentools42 packages. Copy the 4.2 xen.gz
484: to /.
485:
486: Ensure that the contents of /etc/rc.d/xen* are correct. Specifically,
487: they must match the package you just installed and not be left over
488: from some previous installation.
489:
490: Enable the correct set of daemons; see the configuring section above.
491: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
492:
493: Ensure that the domU config files are valid for the new version.
494: Specifically, for 4.x remove autorestart=True, and ensure that disks
495: are specified with numbers as the second argument, as the examples
496: above show, and not NetBSD device names.
497:
498: Hardware known to work
499: ----------------------
500:
501: Arguably, this section is misplaced, and there should be a page of
502: hardware that runs NetBSD/amd64 well, with the mostly-well-founded
503: assumption that NetBSD/xen runs fine on any modern hardware that
504: NetBSD/amd64 runs well on. Until then, we give motherboard/CPU (and
505: sometimes RAM) pairs/triples to aid those choosing a motherboard.
506: Note that Xen systems usually do not run X, so a listing here does not
507: imply that X works at all.
508:
509: Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
510: Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
511: ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
512:
513: Older hardware:
514:
515: Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
516: INTEL DG33FB, "Intel(R) Core(TM)2 Duo CPU E6850 @ 3.00GHz"
517: INTEL DG33FB, "Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz"
518:
519: Running Xen under qemu
520: ----------------------
521:
522: The astute reader will note that this section is somewhat twisted.
523: However, it can be useful to run Xen under qemu either because the
524: version of NetBSD as a dom0 does not run on the hardware in use, or to
525: generate automated test cases involving Xen.
526:
527: In 2015-01, the following combination was reported to mostly work:
528:
529: host OS: NetBSD/amd64 6.1.4
530: qemu: 2.2.0 from pkgsrc
531: Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
532: dom0 kernel: NetBSD/amd64 6.1.5
533: Xen tools: xentools42-4.2.5 from pkgsrc
534:
535: See [PR 47720](https://gnats.netbsd.org/47720) for a problem with dom0
536: shutdown.
537:
538: Unprivileged domains (domU)
539: ===========================
540:
541: This section describes general concepts about domUs. It does not
542: address specific domU operating systems or how to install them. The
543: config files for domUs are typically in /usr/pkg/etc/xen, and are
544: typically named so that the file name, domU name and the domU's host
545: name match.
546:
547: The domU is provided with CPU and memory by Xen, configured by the
548: dom0. The domU is provided with disk and network by the dom0,
549: mediated by Xen, and configured in the dom0.
550:
551: Entropy in domUs can be an issue; physical disks and network are on
552: the dom0. NetBSD's /dev/random system works, but is often challenged.
553:
554: Config files
555: ------------
556:
557: There is no good order to present config files and the concepts
558: surrounding what is being configured. We first show an example config
559: file, and then in the various sections give details.
560:
561: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
562: for a large number of well-commented examples, mostly for running
563: GNU/Linux.
564:
565: The following is an example minimal domain configuration file
566: "/usr/pkg/etc/xen/foo". It is (with only a name change) an actual
567: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
568: i386 domU). The domU serves as a network file server.
569:
570: # -*- mode: python; -*-
571:
572: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
573: memory = 1024
574: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
575: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
576: 'file:/n0/xen/foo-wd1,0x1,w' ]
577:
578: The domain will have the same name as the file. The kernel has the
579: host/domU name in it, so that on the dom0 one can update the various
580: domUs independently. The vif line causes an interface to be provided,
581: with a specific mac address (do not reuse MAC addresses!), in bridge
582: mode. Two disks are provided, and they are both writable; the bits
583: are stored in files and Xen attaches them to a vnd(4) device in the
584: dom0 on domain creation. The system treats xbd0 as the boot device
585: without needing explicit configuration.
586:
587: By default xm looks for domain config files in /usr/pkg/etc/xen. Note
588: that "xm create" takes the name of a config file, while other commands
589: take the name of a domain. To create the domain, connect to the
590: console, create the domain while attaching the console, shutdown the
591: domain, and see if it has finished stopping, do (or xl with Xen >=
592: 4.2):
593:
594: xm create foo
595: xm console foo
596: xm create -c foo
597: xm shutdown foo
598: xm list
599:
600: Typing ^] will exit the console session. Shutting down a domain is
601: equivalent to pushing the power button; a NetBSD domU will receive a
602: power-press event and do a clean shutdown. Shutting down the dom0
603: will trigger controlled shutdowns of all configured domUs.
604:
605: domU kernels
606: ------------
607:
608: On a physical computer, the BIOS reads sector 0, and a chain of boot
609: loaders finds and loads a kernel. Normally this comes from the root
610: file system. With Xen domUs, the process is totally different. The
611: normal path is for the domU kernel to be a file in the dom0's
612: file system. At the request of the dom0, Xen loads that kernel into a
613: new domU instance and starts execution. While domU kernels can be
614: anyplace, reasonable places to store domU kernels on the dom0 are in /
615: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
616: config files), or in /u0/xen (where the vdisks are).
617:
618: Note that loading the domU kernel from the dom0 implies that boot
619: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
620: See the VPS section near the end for discussion of alternate ways to
621: obtain domU kernels.
622:
623: CPU and memory
624: --------------
625:
626: A domain is provided with some number of vcpus, less than the number
627: of CPUs seen by the hypervisor. (For a dom0, this is controlled by
628: the boot argument "dom0_max_vcpus=1".) For a domU, it is controlled
629: from the config file by the "vcpus = N" directive.
630:
631: A domain is provided with memory; this is controlled in the config
632: file by "memory = N" (in megabytes). In the straightforward case, the
633: sum of the the memory allocated to the dom0 and all domUs must be less
634: than the available memory.
635:
636: Xen also provides a "balloon" driver, which can be used to let domains
637: use more memory temporarily. TODO: Explain better, and explain how
638: well it works with NetBSD.
639:
640: Virtual disks
641: -------------
642:
643: With the file/vnd style, typically one creates a directory,
644: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
645: domUs. Then, for each domU disk, one writes zeros to a file that then
646: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
647: for the first virtual disk for the domU called foo. Writing zeros to
648: the file serves two purposes. One is that preallocating the contents
649: improves performance. The other is that vnd on sparse files has
650: failed to work. TODO: give working/notworking NetBSD versions for
651: sparse vnd and gnats reference. Note that the use of file/vnd for Xen
652: is not really different than creating a file-backed virtual disk for
653: some other purpose, except that xentools handles the vnconfig
654: commands. To create an empty 4G virtual disk, simply do
655:
656: dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
657:
658: Do not use qemu-img-xen, because this will create sparse file. There
659: have been recent (2015) reports of sparse vnd(4) devices causing
660: lockups, but there is apparently no PR.
661:
662: With the lvm style, one creates logical devices. They are then used
663: similarly to vnds. TODO: Add an example with lvm.
664:
665: In domU config files, the disks are defined as a sequence of 3-tuples.
666: The first element is "method:/path/to/disk". Common methods are
667: "file:" for file-backed vnd. and "phy:" for something that is already
668: a (TODO: character or block) device.
669:
670: The second element is an artifact of how virtual disks are passed to
671: Linux, and a source of confusion with NetBSD Xen usage. Linux domUs
672: are given a device name to associate with the disk, and values like
673: "hda1" or "sda1" are common. In a NetBSD domU, the first disk appears
674: as xbd0, the second as xbd1, and so on. However, xm/xl demand a
675: second argument. The name given is converted to a major/minor by
676: calling stat(2) on the name in /dev and this is passed to the domU.
677: In the general case, the dom0 and domU can be different operating
678: systems, and it is an unwarranted assumption that they have consistent
679: numbering in /dev, or even that the dom0 OS has a /dev. With NetBSD
680: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
681: for the second works fine and avoids this issue. For a GNU/Linux
682: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
683: /dev/hda1.
684:
685: The third element is "w" for writable disks, and "r" for read-only
686: disks.
687:
688: Note that NetBSD by default creates only vnd[0123]. If you need more
689: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
690: dom0.
691:
692: Note that NetBSD by default creates only xbd[0123]. If you need more
693: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
694:
695: Virtual Networking
696: ------------------
697:
698: Xen provides virtual Ethernets, each of which connects the dom0 and a
699: domU. For each virtual network, there is an interface "xvifN.M" in
700: the dom0, and in domU index N, a matching interface xennetM (NetBSD
701: name). The interfaces behave as if there is an Ethernet with two
702: adapters connected. From this primitive, one can construct various
703: configurations. We focus on two common and useful cases for which
704: there are existing scripts: bridging and NAT.
705:
706: With bridging (in the example above), the domU perceives itself to be
707: on the same network as the dom0. For server virtualization, this is
708: usually best. Bridging is accomplished by creating a bridge(4) device
709: and adding the dom0's physical interface and the various xvifN.0
710: interfaces to the bridge. One specifies "bridge=bridge0" in the domU
711: config file. The bridge must be set up already in the dom0; an
712: example /etc/ifconfig.bridge0 is:
713:
714: create
715: up
716: !brconfig bridge0 add wm0
717:
718: With NAT, the domU perceives itself to be behind a NAT running on the
719: dom0. This is often appropriate when running Xen on a workstation.
720: TODO: NAT appears to be configured by "vif = [ '' ]".
721:
722: The MAC address specified is the one used for the interface in the new
723: domain. The interface in dom0 will use this address XOR'd with
724: 00:00:00:01:00:00. Random MAC addresses are assigned if not given.
725:
726: Sizing domains
727: --------------
728:
729: Modern x86 hardware has vast amounts of resources. However, many
730: virtual servers can function just fine on far less. A system with
731: 512M of RAM and a 4G disk can be a reasonable choice. Note that it is
732: far easier to adjust virtual resources than physical ones. For
733: memory, it's just a config file edit and a reboot. For disk, one can
734: create a new file and vnconfig it (or lvm), and then dump/restore,
735: just like updating physical disks, but without having to be there and
736: without those pesky connectors.
737:
738: Starting domains automatically
739: ------------------------------
740:
741: To start domains foo at bar at boot and shut them down cleanly on dom0
742: shutdown, in rc.conf add:
743:
744: xendomains="foo bar"
745:
746: Note that earlier versions of the xentools41 xendomains rc.d script
747: used xl, when one should use xm with 4.1.
748:
749: Creating specific unprivileged domains (domU)
750: =============================================
751:
752: Creating domUs is almost entirely independent of operating system. We
753: have already presented the basics of config files. Note that you must
754: have already completed the dom0 setup so that "xl list" (or "xm list")
755: works.
756:
757: Creating an unprivileged NetBSD domain (domU)
758: ---------------------------------------------
759:
760: See the earlier config file, and adjust memory. Decide on how much
761: storage you will provide, and prepare it (file or lvm).
762:
763: While the kernel will be obtained from the dom0 file system, the same
764: file should be present in the domU as /netbsd so that tools like
765: savecore(8) can work. (This is helpful but not necessary.)
766:
767: The kernel must be specifically for Xen and for use as a domU. The
768: i386 and amd64 provide the following kernels:
769:
770: i386 XEN3_DOMU
771: i386 XEN3PAE_DOMU
772: amd64 XEN3_DOMU
773:
774: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
775: use the PAE version of the i386 kernel.
776:
777: This will boot NetBSD, but this is not that useful if the disk is
778: empty. One approach is to unpack sets onto the disk outside of xen
779: (by mounting it, just as you would prepare a physical disk for a
780: system you can't run the installer on).
781:
782: A second approach is to run an INSTALL kernel, which has a miniroot
783: and can load sets from the network. To do this, copy the INSTALL
784: kernel to / and change the kernel line in the config file to:
785:
786: kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
787:
788: Then, start the domain as "xl create -c configname".
789:
790: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
791: line should be used in the config file.
792:
793: disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
794:
795: After booting the domain, the option to install via CDROM may be
796: selected. The CDROM device should be changed to `xbd1d`.
797:
798: Once done installing, "halt -p" the new domain (don't reboot or halt,
799: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
800: config file), switch the config file back to the XEN3_DOMU kernel,
801: and start the new domain again. Now it should be able to use "root on
802: xbd0a" and you should have a, functional NetBSD domU.
803:
804: TODO: check if this is still accurate.
805: When the new domain is booting you'll see some warnings about *wscons*
806: and the pseudo-terminals. These can be fixed by editing the files
807: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
808: `/etc/ttys`, except *console*, like this:
809:
810: console "/usr/libexec/getty Pc" vt100 on secure
811: ttyE0 "/usr/libexec/getty Pc" vt220 off secure
812: ttyE1 "/usr/libexec/getty Pc" vt220 off secure
813: ttyE2 "/usr/libexec/getty Pc" vt220 off secure
814: ttyE3 "/usr/libexec/getty Pc" vt220 off secure
815:
816: Finally, all screens must be commented out from `/etc/wscons.conf`.
817:
818: It is also desirable to add
819:
820: powerd=YES
821:
822: in rc.conf. This way, the domain will be properly shut down if
823: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
824:
825: It is not strictly necessary to have a kernel (as /netbsd) in the domU
826: file system. However, various programs (e.g. netstat) will use that
827: kernel to look up symbols to read from kernel virtual memory. If
828: /netbsd is not the running kernel, those lookups will fail. (This is
829: not really a Xen-specific issue, but because the domU kernel is
830: obtained from the dom0, it is far more likely to be out of sync or
831: missing with Xen.)
832:
833: Creating an unprivileged Linux domain (domU)
834: --------------------------------------------
835:
836: Creating unprivileged Linux domains isn't much different from
837: unprivileged NetBSD domains, but there are some details to know.
838:
839: First, the second parameter passed to the disk declaration (the '0x1' in
840: the example below)
841:
842: disk = [ 'phy:/dev/wd0e,0x1,w' ]
843:
844: does matter to Linux. It wants a Linux device number here (e.g. 0x300
845: for hda). Linux builds device numbers as: (major \<\< 8 + minor).
846: So, hda1 which has major 3 and minor 1 on a Linux system will have
847: device number 0x301. Alternatively, devices names can be used (hda,
848: hdb, ...) as xentools has a table to map these names to devices
849: numbers. To export a partition to a Linux guest we can use:
850:
851: disk = [ 'phy:/dev/wd0e,0x300,w' ]
852: root = "/dev/hda1 ro"
853:
854: and it will appear as /dev/hda on the Linux system, and be used as root
855: partition.
856:
857: To install the Linux system on the partition to be exported to the
858: guest domain, the following method can be used: install
859: sysutils/e2fsprogs from pkgsrc. Use mke2fs to format the partition
860: that will be the root partition of your Linux domain, and mount it.
861: Then copy the files from a working Linux system, make adjustments in
862: `/etc` (fstab, network config). It should also be possible to extract
863: binary packages such as .rpm or .deb directly to the mounted partition
864: using the appropriate tool, possibly running under NetBSD's Linux
865: emulation. Once the file system has been populated, umount it. If
866: desirable, the file system can be converted to ext3 using tune2fs -j.
867: It should now be possible to boot the Linux guest domain, using one of
868: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
869:
870: To get the Linux console right, you need to add:
871:
872: extra = "xencons=tty1"
873:
874: to your configuration since not all Linux distributions auto-attach a
875: tty to the xen console.
876:
877: Creating an unprivileged Solaris domain (domU)
878: ----------------------------------------------
879:
880: See possibly outdated
881: [Solaris domU instructions](/ports/xen/howto-solaris/).
882:
883:
884: PCI passthrough: Using PCI devices in guest domains
885: ---------------------------------------------------
886:
887: The dom0 can give other domains access to selected PCI
888: devices. This can allow, for example, a non-privileged domain to have
889: access to a physical network interface or disk controller. However,
890: keep in mind that giving a domain access to a PCI device most likely
891: will give the domain read/write access to the whole physical memory,
892: as PCs don't have an IOMMU to restrict memory access to DMA-capable
893: device. Also, it's not possible to export ISA devices to non-dom0
894: domains, which means that the primary VGA adapter can't be exported.
895: A guest domain trying to access the VGA registers will panic.
896:
897: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
898: not been ported to later versions at this time.
899:
900: For a PCI device to be exported to a domU, is has to be attached to
901: the "pciback" driver in dom0. Devices passed to the dom0 via the
902: pciback.hide boot parameter will attach to "pciback" instead of the
903: usual driver. The list of devices is specified as "(bus:dev.func)",
904: where bus and dev are 2-digit hexadecimal numbers, and func a
905: single-digit number:
906:
907: pciback.hide=(00:0a.0)(00:06.0)
908:
909: pciback devices should show up in the dom0's boot messages, and the
910: devices should be listed in the `/kern/xen/pci` directory.
911:
912: PCI devices to be exported to a domU are listed in the "pci" array of
913: the domU's config file, with the format "0000:bus:dev.func".
914:
915: pci = [ '0000:00:06.0', '0000:00:0a.0' ]
916:
917: In the domU an "xpci" device will show up, to which one or more pci
918: buses will attach. Then the PCI drivers will attach to PCI buses as
919: usual. Note that the default NetBSD DOMU kernels do not have "xpci"
920: or any PCI drivers built in by default; you have to build your own
921: kernel to use PCI devices in a domU. Here's a kernel config example;
922: note that only the "xpci" lines are unusual.
923:
924: include "arch/i386/conf/XEN3_DOMU"
925:
926: # Add support for PCI buses to the XEN3_DOMU kernel
927: xpci* at xenbus ?
928: pci* at xpci ?
929:
930: # PCI USB controllers
931: uhci* at pci? dev ? function ? # Universal Host Controller (Intel)
932:
933: # USB bus support
934: usb* at uhci?
935:
936: # USB Hubs
937: uhub* at usb?
938: uhub* at uhub? port ? configuration ? interface ?
939:
940: # USB Mass Storage
941: umass* at uhub? port ? configuration ? interface ?
942: wd* at umass?
943: # SCSI controllers
944: ahc* at pci? dev ? function ? # Adaptec [23]94x, aic78x0 SCSI
945:
946: # SCSI bus support (for both ahc and umass)
947: scsibus* at scsi?
948:
949: # SCSI devices
950: sd* at scsibus? target ? lun ? # SCSI disk drives
951: cd* at scsibus? target ? lun ? # SCSI CD-ROM drives
952:
953:
954: NetBSD as a domU in a VPS
955: =========================
956:
957: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
958: hardware. This section explains how to deal with Xen in a domU as a
959: virtual private server where you do not control or have access to the
960: dom0. This is not intended to be an exhaustive list of VPS providers;
961: only a few are mentioned that specifically support NetBSD.
962:
963: VPS operators provide varying degrees of access and mechanisms for
964: configuration. The big issue is usually how one controls which kernel
965: is booted, because the kernel is nominally in the dom0 file system (to
966: which VPS users do not normally have access). A second issue is how
967: to install NetBSD.
968: A VPS user may want to compile a kernel for security updates, to run
969: npf, run IPsec, or any other reason why someone would want to change
970: their kernel.
971:
972: One approach is to have an administrative interface to upload a kernel,
973: or to select from a prepopulated list. Other approaches are pygrub
974: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
975: kernel from the domU file system. This is closer to a regular physical
976: computer, where someone who controls a machine can replace the kernel.
977:
978: A second issue is multiple CPUs. With NetBSD 6, domUs support
979: multiple vcpus, and it is typical for VPS providers to enable multiple
980: CPUs for NetBSD domUs.
981:
982: pygrub
983: -------
984:
985: pygrub runs in the dom0 and looks into the domU file system. This
986: implies that the domU must have a kernel in a file system in a format
987: known to pygrub. As of 2014, pygrub seems to be of mostly historical
988: interest.
989:
990: pvgrub
991: ------
992:
993: pvgrub is a version of grub that uses PV operations instead of BIOS
994: calls. It is booted from the dom0 as the domU kernel, and then reads
995: /grub/menu.lst and loads a kernel from the domU file system.
996:
997: [Panix](http://www.panix.com/) lets users use pvgrub. Panix reports
998: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
999: (and hence with defaults from "newfs -O 2"). See [Panix's pvgrub
1000: page](http://www.panix.com/v-colo/grub.html), which describes only
1001: Linux but should be updated to cover NetBSD :-).
1002:
1003: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
1004: their own kernel. See then [prgmr.com NetBSD
1005: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
1006: (which is in need of updating).
1007:
1008: It appears that [grub's FFS
1009: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
1010: does not support all aspects of modern FFS, but there are also reports
1011: that FFSv2 works fine. At prgmr, typically one has an ext2 or FAT
1012: partition for the kernel with the intent that grub can understand it,
1013: which leads to /netbsd not being the actual kernel. One must remember
1014: to update the special boot partition.
1015:
1016: Amazon
1017: ------
1018:
1019: See the [Amazon EC2 page](/amazon_ec2/).
1020:
1021: TODO items for improving NetBSD/xen
1022: ===================================
1023:
1024: * Make the NetBSD dom0 kernel work with SMP.
1025: * Test the Xen 4.5 packages adequately to be able to recommend them as
1026: the standard approach.
1027: * Get PCI passthrough working on Xen 4.5
1028: * Get pvgrub into pkgsrc, either via xentools or separately.
1029: * grub
1030: * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
1031: fragsize/blocksize (UFS2 support may be present; the point is to
1032: make it so that with any UFS1/UFS2 file system setup that works
1033: with NetBSD grub will also work).
1034: See [pkg/40258](https://gnats.netbsd.org/40258).
1035: * Push patches upstream.
1036: * Get UFS2 patches into pvgrub.
1037: * Add support for PV ops to a version of /boot, and make it usable as
1038: a kernel in Xen, similar to pvgrub.
1039:
1040: Random pointers
1041: ===============
1042:
1043: This section contains links from elsewhere not yet integrated into the
1044: HOWTO, and other guides.
1045:
1046: * http://www.lumbercartel.ca/library/xen/
1047: * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html
1048: * https://gmplib.org/~tege/xen.html
CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb