1: Introduction
2: ============
3:
4: [![[Xen
5: screenshot]](http://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](http://www.netbsd.org/gallery/in-Action/hubertf-xen.png)
6:
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.
16:
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.
24:
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.
29:
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.
36:
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.)
40:
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.
46:
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.
51:
52: NetBSD used to support Xen2; this has been removed.
53:
54: Prerequisites
55: -------------
56:
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.
62:
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/).
67:
68: Versions of Xen and NetBSD
69: ==========================
70:
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.
75:
76: Xen
77: ---
78:
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.
83:
84: xenkernel3 and xenkernel33 provide Xen 3.1 and 3.3. These no longer
85: receive security patches and should not be used. Xen 3.1 supports PCI
86: passthrough. Xen 3.1 supports non-PAE on i386.
87:
88: xenkernel41 provides Xen 4.1. This is no longer maintained by Xen,
89: but as of 2014-12 receives backported security patches. It is a
90: reasonable although trailing-edge choice.
91:
92: xenkernel42 provides Xen 4.2. This is maintained by Xen, but old as
93: of 2014-12.
94:
95: xenkernel45 provides Xen 4.5. This is new to pkgsrc as of 2015-01 and
96: not yet recommended for other than experimental/testing use.
97:
98: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
99:
100: Ideally newer versions of Xen will be added to pkgsrc.
101:
102: Note that NetBSD support is called XEN3. It works with Xen 3 and Xen
103: 4 because the hypercall interface has been stable.
104:
105: Xen command program
106: -------------------
107:
108: Early Xen used a program called xm to manipulate the system from the
109: dom0. Starting in 4.1, a replacement program with similar behavior
110: called xl is provided, but it does not work well in 4.1. In 4.2, both
111: xm and xl work fine. 4.4 is the last version that has xm. You must
112: choose one or the other, because it affects which daemons you run.
113:
114: NetBSD
115: ------
116:
117: The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
118: reasonable choices, with more or less the same considerations for
119: non-Xen use. Therefore, netbsd-6 is recommended as the stable version
120: of the most recent release for production use. For those wanting to
121: learn Xen or without production stability concerns, netbsd-7 is likely
122: most appropriate.
123:
124: As of NetBSD 6, a NetBSD domU will support multiple vcpus. There is
125: no SMP support for NetBSD as dom0. (The dom0 itself doesn't really
126: need SMP; the lack of support is really a problem when using a dom0 as
127: a normal computer.)
128:
129: Architecture
130: ------------
131:
132: Xen itself can run on i386 or amd64 machines. (Practically, almost
133: any computer where one would want to run Xen today supports amd64.)
134:
135: Xen, the dom0 kernel, and each domU kernel can be either i386 or
136: amd64. When building a xenkernel package, one obtains i386 on an i386
137: host, and amd64 on an amd64 host. If the xen kernel is i386, then the
138: dom0 kernel and all domU kernels must be i386. With an amd64 xen
139: kernel, an amd64 dom0 kernel is known to work, and an i386 dom0 kernel
140: should in theory work. An amd64 xen/dom0 is known to support both
141: i386 and amd64 domUs.
142:
143: i386 dom0 and domU kernels must be PAE (except for Xen 3.1); these are
144: built by default. (Note that emacs (at least) fails if run on i386
145: with PAE when built without, and vice versa, presumably due to bugs in
146: the undump code.)
147:
148: Because of the above, the standard approach is to use amd64 for the
149: dom0.
150:
151: Xen 4.2 is the last version to support i386 as a host. TODO: Clarify
152: if this is about the CPU, the xen kernel, or the dom0 kernel having to
153: be amd64.
154:
155:
156: Stability
157: ---------
158:
159: Mostly, NetBSD as a dom0 or domU is quite stable.
160: However, there are some open PRs indicating problems.
161:
162: - [PR 48125](http://gnats.netbsd.org/48125)
163: - [PR 47720](http://gnats.netbsd.org/47720)
164:
165: Note also that there are issues with sparse vnd(4) instances, but
166: these are not about Xen.
167:
168: Recommendation
169: --------------
170:
171: Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
172: xl, the NetBSD 6 stable branch, and to use an amd64 kernel as the
173: dom0. Either the i386 or amd64 of NetBSD may be used as domUs.
174:
175: Build problems
176: --------------
177:
178: Ideally, all versions of Xen in pkgsrc would build on all versions of
179: NetBSD on both i386 and amd64. However, that isn't the case. Besides
180: aging code and aging compilers, qemu (included in xentools for HVM
181: support) is difficult to build. The following are known to work or FAIL:
182:
183: xenkernel3 netbsd-5 amd64
184: xentools3 netbsd-5 amd64
185: xentools3=hvm netbsd-5 amd64 ????
186: xenkernel33 netbsd-5 amd64
187: xentools33 netbsd-5 amd64
188: xenkernel41 netbsd-5 amd64
189: xentools41 netbsd-5 amd64
190: xenkernel42 netbsd-5 amd64
191: xentools42 netbsd-5 amd64
192:
193: xenkernel3 netbsd-6 i386 FAIL
194: xentools3 netbsd-6 i386
195: xentools3-hvm netbsd-6 i386 FAIL (dependencies fail)
196: xenkernel33 netbsd-6 i386
197: xentools33 netbsd-6 i386
198: xenkernel41 netbsd-6 i386
199: xentools41 netbsd-6 i386
200: xenkernel42 netbsd-6 i386
201: xentools42 netbsd-6 i386 *MIXED
202:
203: (all 3 and 33 seem to FAIL)
204: xenkernel41 netbsd-7 i386
205: xentools41 netbsd-7 i386
206: xenkernel42 netbsd-7 i386
207: xentools42 netbsd-7 i386 ??FAIL
208:
209: (*On netbsd-6 i386, there is a xentools42 in the 2014Q3 official builds,
210: but it does not build for gdt.)
211:
212: NetBSD as a dom0
213: ================
214:
215: NetBSD can be used as a dom0 and works very well. The following
216: sections address installation, updating NetBSD, and updating Xen.
217: Note that it doesn't make sense to talk about installing a dom0 OS
218: without also installing Xen itself. We first address installing
219: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
220: NetBSD install to a dom0 install by just changing the kernel and boot
221: configuration.
222:
223: For experimenting with Xen, a machine with as little as 1G of RAM and
224: 100G of disk can work. For running many domUs in productions, far
225: more will be needed.
226:
227: Styles of dom0 operation
228: ------------------------
229:
230: There are two basic ways to use Xen. The traditional method is for
231: the dom0 to do absolutely nothing other than providing support to some
232: number of domUs. Such a system was probably installed for the sole
233: purpose of hosting domUs, and sits in a server room on a UPS.
234:
235: The other way is to put Xen under a normal-usage computer, so that the
236: dom0 is what the computer would have been without Xen, perhaps a
237: desktop or laptop. Then, one can run domUs at will. Purists will
238: deride this as less secure than the previous approach, and for a
239: computer whose purpose is to run domUs, they are right. But Xen and a
240: dom0 (without domUs) is not meaningfully less secure than the same
241: things running without Xen. One can boot Xen or boot regular NetBSD
242: alternately with little problems, simply refraining from starting the
243: Xen daemons when not running Xen.
244:
245: Note that NetBSD as dom0 does not support multiple CPUs. This will
246: limit the performance of the Xen/dom0 workstation approach. In theory
247: the only issue is that the "backend drivers" are not yet MPSAFE:
248: http://mail-index.netbsd.org/netbsd-users/2014/08/29/msg015195.html
249:
250: Installation of NetBSD
251: ----------------------
252:
253: First,
254: [install NetBSD/amd64](/guide/inst/)
255: just as you would if you were not using Xen.
256: However, the partitioning approach is very important.
257:
258: If you want to use RAIDframe for the dom0, there are no special issues
259: for Xen. Typically one provides RAID storage for the dom0, and the
260: domU systems are unaware of RAID. The 2nd-stage loader bootxx_* skips
261: over a RAID1 header to find /boot from a filesystem within a RAID
262: partition; this is no different when booting Xen.
263:
264: There are 4 styles of providing backing storage for the virtual disks
265: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
266:
267: With raw partitions, one has a disklabel (or gpt) partition sized for
268: each virtual disk to be used by the domU. (If you are able to predict
269: how domU usage will evolve, please add an explanation to the HOWTO.
270: Seriously, needs tend to change over time.)
271:
272: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
273: for domU disks. This is almost as efficient as raw disk partitions
274: and more flexible. Hence raw disk partitions should typically not
275: be used.
276:
277: One can use files in the dom0 filesystem, typically created by dd'ing
278: /dev/zero to create a specific size. This is somewhat less efficient,
279: but very convenient, as one can cp the files for backup, or move them
280: between dom0 hosts.
281:
282: Finally, in theory one can place the files backing the domU disks in a
283: SAN. (This is an invitation for someone who has done this to add a
284: HOWTO page.)
285:
286: Installation of Xen
287: -------------------
288:
289: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
290: pkgsrc (or another matching pair).
291: See [the pkgsrc
292: documentation](http://www.NetBSD.org/docs/pkgsrc/) for help with pkgsrc.
293:
294: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm. More
295: recent versions have HVM support integrated in the main xentools
296: package. It is entirely reasonable to run only PV guests.
297:
298: Next you need to install the selected Xen kernel itself, which is
299: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz". Copy it to /.
300: For debugging, one may copy xen-debug.gz; this is conceptually similar
301: to DIAGNOSTIC and DEBUG in NetBSD. xen-debug.gz is basically only
302: useful with a serial console. Then, place a NetBSD XEN3_DOM0 kernel
303: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
304: of a NetBSD build. If using i386, use
305: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz. (If using Xen
306: 3.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen. But you
307: should not use Xen 3.1.) Both xen and the NetBSD kernel may be (and
308: typically are) left compressed.
309:
310: In a dom0 kernel, kernfs is mandatory for xend to comunicate with the
311: kernel, so ensure that /kern is in fstab. TODO: Say this is default,
312: or file a PR and give a reference.
313:
314: Because you already installed NetBSD, you have a working boot setup
315: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
316: beginning of your root filesystem, /boot present, and likely
317: /boot.cfg. (If not, fix before continuing!)
318:
319: Add a line to to /boot.cfg to boot Xen. See boot.cfg(5) for an
320: example. The basic line is
321:
322: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
323:
324: which specifies that the dom0 should have 256M, leaving the rest to be
325: allocated for domUs. To use a serial console, use
326:
327: menu=Xen:load /netbsd-XEN3_DOM0.gz console=com0;multiboot /xen.gz dom0_mem=256M console=com1 com1=9600,8n1
328:
329: which will use the first serial port for Xen (which counts starting
330: from 1), forcing speed/parity, and also for NetBSD (which counts
331: starting at 0). In an attempt to add performance, one can also add
332:
333: dom0_max_vcpus=1 dom0_vcpus_pin
334:
335: to force only one vcpu to be provided (since NetBSD dom0 can't use
336: more) and to pin that vcpu to a physical cpu. TODO: benchmark this.
337:
338: Xen has [many boot
339: options](http://xenbits.xenproject.org/docs/4.5-testing/misc/xen-command-line.html),
340: and other tham dom0 memory and max_vcpus, they are generally not
341: necessary.
342:
343: As with non-Xen systems, you should have a line to boot /netbsd (a
344: kernel that works without Xen) and fallback versions of the non-Xen
345: kernel, Xen, and the dom0 kernel.
346:
347: Now, reboot so that you are running a DOM0 kernel under Xen, rather
348: than GENERIC without Xen.
349:
350: Using grub (historic)
351: ---------------------
352:
353: Before NetBSD's native bootloader could support Xen, the use of
354: grub was recommended. If necessary, see the
355: [old grub information](/ports/xen/howto-grub/).
356:
357: The [HowTo on Installing into
358: RAID-1](http://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
359: explains how to set up booting a dom0 with Xen using grub with
360: NetBSD's RAIDframe. (This is obsolete with the use of NetBSD's native
361: boot.)
362:
363: Configuring Xen
364: ---------------
365:
366: Xen logs will be in /var/log/xen.
367:
368: Now, you have a system that will boot Xen and the dom0 kernel, but not
369: do anything else special. Make sure that you have rebooted into Xen.
370: There will be no domUs, and none can be started because you still have
371: to configure the dom0 tools. The daemons which should be run vary
372: with Xen version and with whether one is using xm or xl. Note that
373: xend is for supporting "xm", and should only be used if you plan on
374: using "xm". Do NOT enable xend if you plan on using "xl" as it will
375: cause problems. Running xl without xencommons=YES (and starting it)
376: will result in a hang (so don't do that; follow the HOWTO!).
377:
378: The installation of NetBSD should already have created devices for xen
379: (xencons, xenevt), but if they are not present, create them:
380:
381: cd /dev && sh MAKEDEV xen
382:
383: TODO: Give 3.1 advice (or remove it from pkgsrc).
384:
385: For 3.3 (and thus xm), add to rc.conf (but note that you should have
386: installed 4.1 or 4.2):
387:
388: xend=YES
389: xenbackendd=YES
390:
391: For 4.1 (and thus xm; xl is believed not to work well), add to rc.conf:
392:
393: xencommons=YES
394: xend=YES
395:
396: (If you are using xentools41 from before 2014-12-26, change
397: rc.d/xendomains to use xm rather than xl.)
398:
399: For 4.2 with xm, add to rc.conf
400:
401: xencommons=YES
402: xend=YES
403:
404: For 4.2 with xl, add to rc.conf:
405:
406: xencommons=YES
407: TODO: explain if there is a xend replacement
408:
409: For 4.5 (and thus with xl), add to rc.conf:
410:
411: xencommons=YES
412: TODO: explain if there is a xend replacement
413:
414: TODO: Recommend for/against xen-watchdog.
415:
416: After you have configured the daemons and either started them (in the
417: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
418: available resources, and running domains. An example with xm follows:
419:
420: # xm dmesg
421: [xen's boot info]
422: # xm info
423: [available memory, etc.]
424: # xm list
425: Name Id Mem(MB) CPU State Time(s) Console
426: Domain-0 0 64 0 r---- 58.1
427:
428: With xl, the commands are the same, and the output may be slightly
429: different. TODO: add example output for xl before the xm example,
430: after confirming on 4.2 and resolving the TODO about rc.conf.
431:
432: ### Issues with xencommons
433:
434: xencommons starts xenstored, which stores data on behalf of dom0 and
435: domUs. It does not currently work to stop and start xenstored.
436: Certainly all domUs should be shutdown first, following the sort order
437: of the rc.d scripts. However, the dom0 sets up state with xenstored,
438: and is not notified when xenstored exits, leading to not recreating
439: the state when the new xenstored starts. Until there's a mechanism to
440: make this work, one should not expect to be able to restart xenstored
441: (and thus xencommons). There is currently no reason to expect that
442: this will get fixed any time soon.
443:
444: anita (for testing NetBSD)
445: --------------------------
446:
447: With the setup so far (assuming 4.2/xl), one should be able to run
448: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
449: root, because anita must create a domU):
450:
451: anita --vmm=xl test file:///usr/obj/i386/
452:
453: Alternatively, one can use --vmm=xm to use xm-based domU creation
454: instead (and must, on Xen <= 4.1). TODO: confirm that anita xl really works.
455:
456: Xen-specific NetBSD issues
457: --------------------------
458:
459: There are (at least) two additional things different about NetBSD as a
460: dom0 kernel compared to hardware.
461:
462: One is that modules are not usable in DOM0 kernels, so one must
463: compile in what's needed. It's not really that modules cannot work,
464: but that modules must be built for XEN3_DOM0 because some of the
465: defines change and the normal module builds don't do this. Basically,
466: enabling Xen changes the kernel ABI, and the module build system
467: doesn't cope with this.
468:
469: The other difference is that XEN3_DOM0 does not have exactly the same
470: options as GENERIC. While it is debatable whether or not this is a
471: bug, users should be aware of this and can simply add missing config
472: items if desired.
473:
474: Updating NetBSD in a dom0
475: -------------------------
476:
477: This is just like updating NetBSD on bare hardware, assuming the new
478: version supports the version of Xen you are running. Generally, one
479: replaces the kernel and reboots, and then overlays userland binaries
480: and adjusts /etc.
481:
482: Note that one must update both the non-Xen kernel typically used for
483: rescue purposes and the DOM0 kernel used with Xen.
484:
485: Converting from grub to /boot
486: -----------------------------
487:
488: These instructions were [TODO: will be] used to convert a system from
489: grub to /boot. The system was originally installed in February of
490: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
491: over time. Before these commands, it was running NetBSD 6 i386, Xen
492: 4.1 and grub, much like the message linked earlier in the grub
493: section.
494:
495: # Install mbr bootblocks on both disks.
496: fdisk -i /dev/rwd0d
497: fdisk -i /dev/rwd1d
498: # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
499: installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
500: installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
501: # Install secondary boot loader
502: cp -p /usr/mdec/boot /
503: # Create boog.cfg following earlier guidance:
504: menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
505: menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
506: menu=GENERIC:boot
507: menu=GENERIC single-user:boot -s
508: menu=GENERIC.ok:boot netbsd.ok
509: menu=GENERIC.ok single-user:boot netbsd.ok -s
510: menu=Drop to boot prompt:prompt
511: default=1
512: timeout=30
513:
514: TODO: actually do this and fix it if necessary.
515:
516: Updating Xen versions
517: ---------------------
518:
519: Updating Xen is conceptually not difficult, but can run into all the
520: issues found when installing Xen. Assuming migration from 4.1 to 4.2,
521: remove the xenkernel41 and xentools41 packages and install the
522: xenkernel42 and xentools42 packages. Copy the 4.2 xen.gz to /.
523:
524: Ensure that the contents of /etc/rc.d/xen* are correct. Enable the
525: correct set of daemons. Ensure that the domU config files are valid
526: for the new version.
527:
528: Hardware known to work
529: ----------------------
530:
531: Arguably, this section is misplaced, and there should be a page of
532: hardware that runs NetBSD/amd64 well, with the mostly-well-founded
533: assumption that NetBSD/xen runs fine on any modern hardware that
534: NetBSD/amd64 runs well on. Until then, we give motherboard/CPU/RAM
535: triples to aid those choosing a motherboard. Note that Xen systems
536: usually do not run X, so a listing here does not imply that X works at
537: all.
538:
539: Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
540: Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
541: ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
542:
543: Older hardware:
544:
545: Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
546:
547: Running Xen under qemu
548: ----------------------
549:
550: The astute reader will note that this section is somewhat twisted.
551: However, it can be useful to run Xen under qemu either because the
552: version of NetBSD as a dom0 does not run on the hardware in use, or to
553: generate automated test cases involving Xen.
554:
555: In 2015-01, the following combination was reported to mostly work:
556:
557: host OS: NetBSD/amd64 6.1.4
558: qemu: 2.2.0 from pkgsrc
559: Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
560: dom0 kernel: NetBSD/amd64 6.1.5
561: Xen tools: xentools42-4.2.5 from pkgsrc
562:
563: See [PR 47720](http://gnats.netbsd.org/47720) for a problem with dom0
564: shutdown.
565:
566: Unprivileged domains (domU)
567: ===========================
568:
569: This section describes general concepts about domUs. It does not
570: address specific domU operating systems or how to install them. The
571: config files for domUs are typically in /usr/pkg/etc/xen, and are
572: typically named so that the file name, domU name and the domU's host
573: name match.
574:
575: The domU is provided with cpu and memory by Xen, configured by the
576: dom0. The domU is provided with disk and network by the dom0,
577: mediated by Xen, and configured in the dom0.
578:
579: Entropy in domUs can be an issue; physical disks and network are on
580: the dom0. NetBSD's /dev/random system works, but is often challenged.
581:
582: Config files
583: ------------
584:
585: There is no good order to present config files and the concepts
586: surrounding what is being configured. We first show an example config
587: file, and then in the various sections give details.
588:
589: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
590: for a large number of well-commented examples, mostly for running
591: GNU/Linux.
592:
593: The following is an example minimal domain configuration file
594: "/usr/pkg/etc/xen/foo". It is (with only a name change) an actual
595: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
596: i386 domU). The domU serves as a network file server.
597:
598: # -*- mode: python; -*-
599:
600: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
601: memory = 1024
602: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
603: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
604: 'file:/n0/xen/foo-wd1,0x1,w' ]
605:
606: The domain will have the same name as the file. The kernel has the
607: host/domU name in it, so that on the dom0 one can update the various
608: domUs independently. The vif line causes an interface to be provided,
609: with a specific mac address (do not reuse MAC addresses!), in bridge
610: mode. Two disks are provided, and they are both writable; the bits
611: are stored in files and Xen attaches them to a vnd(4) device in the
612: dom0 on domain creation. The system treates xbd0 as the boot device
613: without needing explicit configuration.
614:
615: By default xm looks for domain config files in /usr/pkg/etc/xen. Note
616: that "xm create" takes the name of a config file, while other commands
617: take the name of a domain. To create the domain, connect to the
618: console, create the domain while attaching the console, shutdown the
619: domain, and see if it has finished stopping, do (or xl with Xen >=
620: 4.2):
621:
622: xm create foo
623: xm console foo
624: xm create -c foo
625: xm shutdown foo
626: xm list
627:
628: Typing ^] will exit the console session. Shutting down a domain is
629: equivalent to pushing the power button; a NetBSD domU will receive a
630: power-press event and do a clean shutdown. Shutting down the dom0
631: will trigger controlled shutdowns of all configured domUs.
632:
633: domU kernels
634: ------------
635:
636: On a physical computer, the BIOS reads sector 0, and a chain of boot
637: loaders finds and loads a kernel. Normally this comes from the root
638: filesystem. With Xen domUs, the process is totally different. The
639: normal path is for the domU kernel to be a file in the dom0's
640: filesystem. At the request of the dom0, Xen loads that kernel into a
641: new domU instance and starts execution. While domU kernels can be
642: anyplace, reasonable places to store domU kernels on the dom0 are in /
643: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
644: config files), or in /u0/xen (where the vdisks are).
645:
646: Note that loading the domU kernel from the dom0 implies that boot
647: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
648: See the VPS section near the end for discussion of alternate ways to
649: obtain domU kernels.
650:
651: CPU and memory
652: --------------
653:
654: A domain is provided with some number of vcpus, less than the number
655: of cpus seen by the hypervisor. (For a dom0, this is controlled by
656: the boot argument "dom0_max_vcpus=1".) For a domU, it is controlled
657: from the config file by the "vcpus = N" directive.
658:
659: A domain is provided with memory; this is controlled in the config
660: file by "memory = N" (in megabytes). In the straightforward case, the
661: sum of the the memory allocated to the dom0 and all domUs must be less
662: than the available memory.
663:
664: Xen also provides a "balloon" driver, which can be used to let domains
665: use more memory temporarily. TODO: Explain better, and explain how
666: well it works with NetBSD.
667:
668: Virtual disks
669: -------------
670:
671: With the file/vnd style, typically one creates a directory,
672: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
673: domUs. Then, for each domU disk, one writes zeros to a file that then
674: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
675: for the first virtual disk for the domU called foo. Writing zeros to
676: the file serves two purposes. One is that preallocating the contents
677: improves performance. The other is that vnd on sparse files has
678: failed to work. TODO: give working/notworking NetBSD versions for
679: sparse vnd. Note that the use of file/vnd for Xen is not really
680: different than creating a file-backed virtual disk for some other
681: purpose, except that xentools handles the vnconfig commands. To
682: create an empty 4G virtual disk, simply do
683:
684: dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
685:
686: Do not use qemu-img-xen, because this will create sparse file. There
687: have been recent (2015) reports of sparse vnd(4) devices causing
688: lockups, but there is apparently no PR.
689:
690: With the lvm style, one creates logical devices. They are then used
691: similarly to vnds. TODO: Add an example with lvm.
692:
693: In domU config files, the disks are defined as a sequence of 3-tuples.
694: The first element is "method:/path/to/disk". Common methods are
695: "file:" for file-backed vnd. and "phy:" for something that is already
696: a (TODO: character or block) device.
697:
698: The second element is an artifact of how virtual disks are passed to
699: Linux, and a source of confusion with NetBSD Xen usage. Linux domUs
700: are given a device name to associate with the disk, and values like
701: "hda1" or "sda1" are common. In a NetBSD domU, the first disk appears
702: as xbd0, the second as xbd1, and so on. However, xm/xl demand a
703: second argument. The name given is converted to a major/minor by
704: calling stat(2) on the name in /dev and this is passed to the domU.
705: In the general case, the dom0 and domU can be different operating
706: systems, and it is an unwarranted assumption that they have consistent
707: numbering in /dev, or even that the dom0 OS has a /dev. With NetBSD
708: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
709: for the second works fine and avoids this issue. For a GNU/Linux
710: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
711: /dev/hda1.
712:
713: The third element is "w" for writable disks, and "r" for read-only
714: disks.
715:
716: Virtual Networking
717: ------------------
718:
719: Xen provides virtual ethernets, each of which connects the dom0 and a
720: domU. For each virtual network, there is an interface "xvifN.M" in
721: the dom0, and in domU index N, a matching interface xennetM (NetBSD
722: name). The interfaces behave as if there is an Ethernet with two
723: adaptors connected. From this primitive, one can construct various
724: configurations. We focus on two common and useful cases for which
725: there are existing scripts: bridging and NAT.
726:
727: With bridging (in the example above), the domU perceives itself to be
728: on the same network as the dom0. For server virtualization, this is
729: usually best. Bridging is accomplished by creating a bridge(4) device
730: and adding the dom0's physical interface and the various xvifN.0
731: interfaces to the bridge. One specifies "bridge=bridge0" in the domU
732: config file. The bridge must be set up already in the dom0; an
733: example /etc/ifconfig.bridge0 is:
734:
735: create
736: up
737: !brconfig bridge0 add wm0
738:
739: With NAT, the domU perceives itself to be behind a NAT running on the
740: dom0. This is often appropriate when running Xen on a workstation.
741: TODO: NAT appears to be configured by "vif = [ '' ]".
742:
743: The MAC address specified is the one used for the interface in the new
744: domain. The interface in dom0 will use this address XOR'd with
745: 00:00:00:01:00:00. Random MAC addresses are assigned if not given.
746:
747: Sizing domains
748: --------------
749:
750: Modern x86 hardware has vast amounts of resources. However, many
751: virtual servers can function just fine on far less. A system with
752: 256M of RAM and a 4G disk can be a reasonable choice. Note that it is
753: far easier to adjust virtual resources than physical ones. For
754: memory, it's just a config file edit and a reboot. For disk, one can
755: create a new file and vnconfig it (or lvm), and then dump/restore,
756: just like updating physical disks, but without having to be there and
757: without those pesky connectors.
758:
759: Starting domains automatically
760: ------------------------------
761:
762: To start domains foo at bar at boot and shut them down cleanly on dom0
763: shutdown, in rc.conf add:
764:
765: xendomains="foo bar"
766:
767: Note that earlier versions of the xentools41 xendomains rc.d scripth
768: usd xl, when one should use xm with 4.1.
769:
770: Creating specific unprivileged domains (domU)
771: =============================================
772:
773: Creating domUs is almost entirely independent of operating system. We
774: have already presented the basics of config files. Note that you must
775: have already completed the dom0 setup so that "xl list" (or "xm list")
776: works.
777:
778: Creating an unprivileged NetBSD domain (domU)
779: ---------------------------------------------
780:
781: See the earlier config file, and adjust memory. Decide on how much
782: storage you will provide, and prepare it (file or lvm).
783:
784: While the kernel will be obtained from the dom0 filesystem, the same
785: file should be present in the domU as /netbsd so that tools like
786: savecore(8) can work. (This is helpful but not necessary.)
787:
788: The kernel must be specifically for Xen and for use as a domU. The
789: i386 and amd64 provide the following kernels:
790:
791: i386 XEN3_DOMU
792: i386 XEN3PAE_DOMU
793: amd64 XEN3_DOMU
794:
795: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
796: use the PAE version of the i386 kernel.
797:
798: This will boot NetBSD, but this is not that useful if the disk is
799: empty. One approach is to unpack sets onto the disk outside of xen
800: (by mounting it, just as you would prepare a physical disk for a
801: system you can't run the installer on).
802:
803: A second approach is to run an INSTALL kernel, which has a miniroot
804: and can load sets from the network. To do this, copy the INSTALL
805: kernel to / and change the kernel line in the config file to:
806:
807: kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
808:
809: Then, start the domain as "xl create -c configname".
810:
811: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
812: line should be used in the config file.
813:
814: disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
815:
816: After booting the domain, the option to install via CDROM may be
817: selected. The CDROM device should be changed to `xbd1d`.
818:
819: Once done installing, "halt -p" the new domain (don't reboot or halt,
820: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
821: config file), switch the config file back to the XEN3_DOMU kernel,
822: and start the new domain again. Now it should be able to use "root on
823: xbd0a" and you should have a, functional NetBSD domU.
824:
825: TODO: check if this is still accurate.
826: When the new domain is booting you'll see some warnings about *wscons*
827: and the pseudo-terminals. These can be fixed by editing the files
828: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
829: `/etc/ttys`, except *console*, like this:
830:
831: console "/usr/libexec/getty Pc" vt100 on secure
832: ttyE0 "/usr/libexec/getty Pc" vt220 off secure
833: ttyE1 "/usr/libexec/getty Pc" vt220 off secure
834: ttyE2 "/usr/libexec/getty Pc" vt220 off secure
835: ttyE3 "/usr/libexec/getty Pc" vt220 off secure
836:
837: Finally, all screens must be commented out from `/etc/wscons.conf`.
838:
839: It is also desirable to add
840:
841: powerd=YES
842:
843: in rc.conf. This way, the domain will be properly shut down if
844: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
845:
846: It is not strictly necessary to have a kernel (as /netbsd) in the domU
847: filesystem. However, various programs (e.g. netstat) will use that
848: kernel to look up symbols to read from kernel virtual memory. If
849: /netbsd is not the running kernel, those lookups will fail. (This is
850: not really a Xen-specific issue, but because the domU kernel is
851: obtained from the dom0, it is far more likely to be out of sync or
852: missing with Xen.)
853:
854: Creating an unprivileged Linux domain (domU)
855: --------------------------------------------
856:
857: Creating unprivileged Linux domains isn't much different from
858: unprivileged NetBSD domains, but there are some details to know.
859:
860: First, the second parameter passed to the disk declaration (the '0x1' in
861: the example below)
862:
863: disk = [ 'phy:/dev/wd0e,0x1,w' ]
864:
865: does matter to Linux. It wants a Linux device number here (e.g. 0x300
866: for hda). Linux builds device numbers as: (major \<\< 8 + minor).
867: So, hda1 which has major 3 and minor 1 on a Linux system will have
868: device number 0x301. Alternatively, devices names can be used (hda,
869: hdb, ...) as xentools has a table to map these names to devices
870: numbers. To export a partition to a Linux guest we can use:
871:
872: disk = [ 'phy:/dev/wd0e,0x300,w' ]
873: root = "/dev/hda1 ro"
874:
875: and it will appear as /dev/hda on the Linux system, and be used as root
876: partition.
877:
878: To install the Linux system on the partition to be exported to the
879: guest domain, the following method can be used: install
880: sysutils/e2fsprogs from pkgsrc. Use mke2fs to format the partition
881: that will be the root partition of your Linux domain, and mount it.
882: Then copy the files from a working Linux system, make adjustments in
883: `/etc` (fstab, network config). It should also be possible to extract
884: binary packages such as .rpm or .deb directly to the mounted partition
885: using the appropriate tool, possibly running under NetBSD's Linux
886: emulation. Once the filesystem has been populated, umount it. If
887: desirable, the filesystem can be converted to ext3 using tune2fs -j.
888: It should now be possible to boot the Linux guest domain, using one of
889: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
890:
891: To get the linux console right, you need to add:
892:
893: extra = "xencons=tty1"
894:
895: to your configuration since not all linux distributions auto-attach a
896: tty to the xen console.
897:
898: Creating an unprivileged Solaris domain (domU)
899: ----------------------------------------------
900:
901: See possibly outdated
902: [Solaris domU instructions](/ports/xen/howto-solaris/).
903:
904:
905: PCI passthrough: Using PCI devices in guest domains
906: ---------------------------------------------------
907:
908: The dom0 can give other domains access to selected PCI
909: devices. This can allow, for example, a non-privileged domain to have
910: access to a physical network interface or disk controller. However,
911: keep in mind that giving a domain access to a PCI device most likely
912: will give the domain read/write access to the whole physical memory,
913: as PCs don't have an IOMMU to restrict memory access to DMA-capable
914: device. Also, it's not possible to export ISA devices to non-dom0
915: domains, which means that the primary VGA adapter can't be exported.
916: A guest domain trying to access the VGA registers will panic.
917:
918: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
919: not been ported to later versions at this time.
920:
921: For a PCI device to be exported to a domU, is has to be attached to
922: the "pciback" driver in dom0. Devices passed to the dom0 via the
923: pciback.hide boot parameter will attach to "pciback" instead of the
924: usual driver. The list of devices is specified as "(bus:dev.func)",
925: where bus and dev are 2-digit hexadecimal numbers, and func a
926: single-digit number:
927:
928: pciback.hide=(00:0a.0)(00:06.0)
929:
930: pciback devices should show up in the dom0's boot messages, and the
931: devices should be listed in the `/kern/xen/pci` directory.
932:
933: PCI devices to be exported to a domU are listed in the "pci" array of
934: the domU's config file, with the format "0000:bus:dev.func".
935:
936: pci = [ '0000:00:06.0', '0000:00:0a.0' ]
937:
938: In the domU an "xpci" device will show up, to which one or more pci
939: busses will attach. Then the PCI drivers will attach to PCI busses as
940: usual. Note that the default NetBSD DOMU kernels do not have "xpci"
941: or any PCI drivers built in by default; you have to build your own
942: kernel to use PCI devices in a domU. Here's a kernel config example;
943: note that only the "xpci" lines are unusual.
944:
945: include "arch/i386/conf/XEN3_DOMU"
946:
947: # Add support for PCI busses to the XEN3_DOMU kernel
948: xpci* at xenbus ?
949: pci* at xpci ?
950:
951: # PCI USB controllers
952: uhci* at pci? dev ? function ? # Universal Host Controller (Intel)
953:
954: # USB bus support
955: usb* at uhci?
956:
957: # USB Hubs
958: uhub* at usb?
959: uhub* at uhub? port ? configuration ? interface ?
960:
961: # USB Mass Storage
962: umass* at uhub? port ? configuration ? interface ?
963: wd* at umass?
964: # SCSI controllers
965: ahc* at pci? dev ? function ? # Adaptec [23]94x, aic78x0 SCSI
966:
967: # SCSI bus support (for both ahc and umass)
968: scsibus* at scsi?
969:
970: # SCSI devices
971: sd* at scsibus? target ? lun ? # SCSI disk drives
972: cd* at scsibus? target ? lun ? # SCSI CD-ROM drives
973:
974:
975: NetBSD as a domU in a VPS
976: =========================
977:
978: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
979: hardware. This section explains how to deal with Xen in a domU as a
980: virtual private server where you do not control or have access to the
981: dom0. This is not intended to be an exhaustive list of VPS providers;
982: only a few are mentioned that specifically support NetBSD.
983:
984: VPS operators provide varying degrees of access and mechanisms for
985: configuration. The big issue is usually how one controls which kernel
986: is booted, because the kernel is nominally in the dom0 filesystem (to
987: which VPS users do not normally have acesss). A second issue is how
988: to install NetBSD.
989: A VPS user may want to compile a kernel for security updates, to run
990: npf, run IPsec, or any other reason why someone would want to change
991: their kernel.
992:
993: One approach is to have an adminstrative interface to upload a kernel,
994: or to select from a prepopulated list. Other approaches are pygrub
995: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
996: kernel from the domU filesystem. This is closer to a regular physical
997: computer, where someone who controls a machine can replace the kernel.
998:
999: A second issue is multiple CPUs. With NetBSD 6, domUs support
1000: multiple vcpus, and it is typical for VPS providers to enable multiple
1001: CPUs for NetBSD domUs.
1002:
1003: pygrub
1004: -------
1005:
1006: pygrub runs in the dom0 and looks into the domU filesystem. This
1007: implies that the domU must have a kernel in a filesystem in a format
1008: known to pygrub. As of 2014, pygrub seems to be of mostly historical
1009: interest.
1010:
1011: pvgrub
1012: ------
1013:
1014: pvgrub is a version of grub that uses PV operations instead of BIOS
1015: calls. It is booted from the dom0 as the domU kernel, and then reads
1016: /grub/menu.lst and loads a kernel from the domU filesystem.
1017:
1018: [Panix](http://www.panix.com/) lets users use pvgrub. Panix reports
1019: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
1020: (and hence with defaults from "newfs -O 2"). See [Panix's pvgrub
1021: page](http://www.panix.com/v-colo/grub.html), which describes only
1022: Linux but should be updated to cover NetBSD :-).
1023:
1024: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
1025: their own kernel. See then [prgmr.com NetBSD
1026: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
1027: (which is in need of updating).
1028:
1029: It appears that [grub's FFS
1030: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
1031: does not support all aspects of modern FFS, but there are also reports
1032: that FFSv2 works fine. At prgmr, typically one has an ext2 or FAT
1033: partition for the kernel with the intent that grub can understand it,
1034: which leads to /netbsd not being the actual kernel. One must remember
1035: to update the special boot partiion.
1036:
1037: Amazon
1038: ------
1039:
1040: See the [Amazon EC2 page](../amazon_ec2/).
1041:
1042: Using npf
1043: ---------
1044:
1045: In standard kernels, npf is a module, and thus cannot be loaded in a
1046: DOMU kernel.
1047:
1048: TODO: Explain how to compile npf into a custom kernel, answering (but
1049: note that the problem was caused by not booting the right kernel)
1050: [this email to
1051: netbsd-users](http://mail-index.netbsd.org/netbsd-users/2014/12/26/msg015576.html).
1052:
1053: TODO items for improving NetBSD/xen
1054: ===================================
1055:
1056: * Make the NetBSD dom0 kernel work with SMP.
1057: * Test the Xen 4.5 packages adequately to be able to recommend them as
1058: the standard approach.
1059: * Get PCI passthrough working on Xen 4.5
1060: * Get pvgrub into pkgsrc, either via xentools or separately.
1061: * grub
1062: * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
1063: fragsize/blocksize (UFS2 support may be present; the point is to
1064: make it so that with any UFS1/UFS2 filesystem setup that works
1065: with NetBSD grub will also work).
1066: See [pkg/40258](http://gnats.netbsd.org/40258).
1067: * Push patches upstream.
1068: * Get UFS2 patches into pvgrub.
1069: * Add support for PV ops to a version of /boot, and make it usable as
1070: a kernel in Xen, similar to pvgrub.
1071: * Solve somehow the issue with modules for GENERIC not being loadable
1072: in a Xen dom0 or domU kernel.
1073:
1074: Random pointers
1075: ===============
1076:
1077: TODO: This section contains links from elsewhere not yet integrated
1078: into the HOWTO.
1079:
1080: * http://www.lumbercartel.ca/library/xen/
1081: * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html
CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb