Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.186
1.144 maxv 1: [[!meta title="Xen HowTo"]]
2:
1.157 maxv 3: Xen is a Type 1 hypervisor which supports running multiple guest operating
4: systems on a single physical machine. One uses the Xen kernel to control the
5: CPU, memory and console, a dom0 operating system which mediates access to
6: other hardware (e.g., disks, network, USB), and one or more domU operating
7: systems which operate in an unprivileged virtualized environment. IO requests
8: from the domU systems are forwarded by the Xen hypervisor to the dom0 to be
9: fulfilled.
1.12 gdt 10:
1.182 gdt 11: This HOWTO presumes a basic familiarity with the Xen system
12: architecture, with installing NetBSD on amd64 hardware, and with
13: installing software from pkgsrc. See also the [Xen
14: website](http://www.xenproject.org/).
15:
16: [[!toc]]
17:
1.183 gdt 18: # Overview
1.182 gdt 19:
20: The basic concept of Xen is that the hypervisor (xenkernel) runs on
21: the hardware, and runs a privileged domain ("dom0") that can access
22: disks/networking/etc. One then runs additonal unprivileged domains
23: (each a "domU"), presumably to do something useful.
24:
25: This HOWTO addresses how to run a NetBSD dom0 (and hence also build
26: xen itself). It also addresses how to run domUs in that environment,
27: and how to deal with having a domU in a Xen environment run by someone
28: else and/or not running NetBSD.
29:
1.186 ! gdt 30: There are many choices one can make; the HOWTO recommends the standard
! 31: approach and limits discussion of alternatives in many cases.
! 32:
1.182 gdt 33: ## Guest Styles
34:
35: Xen supports different styles of guests.
1.149 maxv 36:
37: [[!table data="""
38: Style of guest |Supported by NetBSD
1.150 maxv 39: PV |Yes (dom0, domU)
40: HVM |Yes (domU)
1.171 gdt 41: PVHVM |current-only (domU)
1.180 gdt 42: PVH |current-only (domU, dom0 not yet)
1.149 maxv 43: """]]
44:
45: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
46: hardware directly, but instead makes hypercalls to the hypervisor; PV
1.170 gdt 47: guests must be specifically coded for Xen.
1.182 gdt 48: See [PV](https://wiki.xen.org/wiki/Paravirtualization_(PV\)).
1.170 gdt 49:
50: In HVM mode, no guest modification is required; however, hardware
51: support is required, such as VT-x on Intel CPUs and SVM on AMD CPUs.
1.177 gdt 52: The dom0 runs qemu to emulate hardware.
1.170 gdt 53:
1.176 gdt 54: In PVHVM mode, the guest runs as HVM, but additionally can use PV
55: drivers for efficiency.
1.182 gdt 56: See [PV on HVM](https://wiki.xen.org/wiki/PV_on_HVM).
1.170 gdt 57:
1.180 gdt 58: There have been two PVH modes: original PVH and PVHv2. Original PVH
59: was based on PV mode and is no longer relevant at all. PVHv2 is
60: basically lightweight HVM with PV drivers. A critical feature of it
61: is that qemu is not needed; the hypervisor can do the emulation that
62: is required. Thus, a dom0 can be PVHv2.
1.182 gdt 63: The source code uses PVH and config files use pvh; this refers to PVHv2.
64: See [PVH(v2)](https://wiki.xenproject.org/wiki/PVH_(v2\)_Domu).
1.180 gdt 65:
1.182 gdt 66: At system boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12 gdt 67: The dom0 can start one or more domUs. (Booting is explained in detail
68: in the dom0 section.)
69:
1.182 gdt 70: ## CPU Architecture
71:
72: Xen runs on x86_64 hardware (the NetBSD amd64 port).
73:
74: There is a concept of Xen running on ARM, but there are no reports of this working with NetBSD.
75:
76: The dom0 system should be amd64. (Instructions for i386PAE dom0 have been removed from the HOWTO.)
1.1 mspo 77:
1.182 gdt 78: The domU can be i386PAE or amd64.
79: i386PAE at one point was considered as [faster](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html) than amd64.
1.150 maxv 80:
1.182 gdt 81: ## Xen Versions
1.15 gdt 82:
1.111 wiz 83: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
1.15 gdt 84: xenkernel and xentools. We will refer only to the kernel versions,
85: but note that both packages must be installed together and must have
86: matching versions.
87:
1.145 maxv 88: Versions available in pkgsrc:
1.85 gdt 89:
1.145 maxv 90: [[!table data="""
1.184 gdt 91: Xen Version |Package Name |Xen CPU Support |EOL'ed By Upstream
92: 4.11 |xenkernel411 |x86_64 |No
93: 4.13 |xenkernel413 |x86_64 |No
1.145 maxv 94: """]]
1.113 gdt 95:
1.96 gdt 96: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
97:
1.166 gdt 98: Older Xen had a python-based management tool called xm, now replaced
1.176 gdt 99: by xl.
1.166 gdt 100:
1.183 gdt 101: ## NetBSD versions
102:
103: Xen has been supported in NetBSD for a long time, at least since 2005.
104: Initially Xen was PV only.
105:
106: NetBSD 8 and up support PV and HVM modes.
107:
108: Support for PVHVM and PVH is available only in NetBSD-current.
109:
1.185 gdt 110: NetBSD as a dom0 does not run SMP, because some drivers are not yet
111: safe for this. \todo Link to more information about what needs work.
112:
113: NetBSD, when run as a domU, can and does typically run SMP.
114:
115: Note: NetBSD support is called XEN3. However, it does support Xen 4,
116: because the hypercall interface has remained identical.
117:
1.166 gdt 118: # Creating a dom0
1.15 gdt 119:
1.158 maxv 120: In order to install a NetBSD as a dom0, one must first install a normal
121: NetBSD system, and then pivot the install to a dom0 install by changing
122: the kernel and boot configuration.
1.15 gdt 123:
1.142 gdt 124: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
125: 512M it worked reliably. This does not make sense, but if you see
126: "not ELF" after Xen boots, try increasing dom0 RAM.
127:
1.181 gdt 128: ## Installation of NetBSD
1.13 gdt 129:
1.157 maxv 130: [Install NetBSD/amd64](/guide/inst/)
1.19 gdt 131: just as you would if you were not using Xen.
1.186 ! gdt 132: Therefore, use the most recent release, or a build from the most recent stable branch.
1.1 mspo 133:
1.181 gdt 134: ## Installation of Xen
1.1 mspo 135:
1.186 ! gdt 136: Use the most recent version of Xen in pkgsrc, unless the DESCR says that it ss not suitable.
! 137: Therefore, choose 4.13.
! 138: In the dom0, install xenkernel413 and xentools413 from pkgsrc.
1.155 maxv 139:
1.186 ! gdt 140: Once this is done, copy the Xen kernel from where pkgsrc puts it to
! 141: where the boot process will be able to find it:
1.155 maxv 142:
143: [[!template id=programlisting text="""
1.186 ! gdt 144: # cp -p /usr/pkg/xen413-kernel/xen.gz /
1.155 maxv 145: """]]
146:
1.159 maxv 147: Then, place a NetBSD XEN3_DOM0 kernel in the `/` directory. Such kernel
148: can either be compiled manually, or downloaded from the NetBSD FTP, for
149: example at:
150:
151: [[!template id=programlisting text="""
1.186 ! gdt 152: ftp.netbsd.org/pub/NetBSD/NetBSD-9.1/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
1.159 maxv 153: """]]
1.20 gdt 154:
1.155 maxv 155: Add a line to /boot.cfg to boot Xen:
1.20 gdt 156:
1.157 maxv 157: [[!template id=filecontent name="/boot.cfg" text="""
1.152 maxv 158: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
159: """]]
1.20 gdt 160:
1.159 maxv 161: This specifies that the dom0 should have 512MB of ram, leaving the rest
162: to be allocated for domUs. To use a serial console, use:
1.77 gdt 163:
1.157 maxv 164: [[!template id=filecontent name="/boot.cfg" text="""
1.152 maxv 165: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
166: """]]
1.77 gdt 167:
168: which will use the first serial port for Xen (which counts starting
1.139 gson 169: from 1, unlike NetBSD which counts starting from 0), forcing
170: speed/parity. Because the NetBSD command line lacks a
171: "console=pc" argument, it will use the default "xencons" console device,
172: which directs the console I/O through Xen to the same console device Xen
173: itself uses (in this case, the serial port).
174:
1.157 maxv 175: In an attempt to add performance, one can also add `dom0_max_vcpus=1 dom0_vcpus_pin`,
1.37 gdt 176: to force only one vcpu to be provided (since NetBSD dom0 can't use
1.158 maxv 177: more) and to pin that vcpu to a physical CPU. Xen has
1.178 gdt 178: [many boot options](http://xenbits.xenproject.org/docs/4.13-testing/misc/xen-command-line.html),
1.111 wiz 179: and other than dom0 memory and max_vcpus, they are generally not
1.93 gdt 180: necessary.
181:
1.186 ! gdt 182: Ensure that the boot scripts installed in
! 183: `/usr/pkg/share/examples/rc.d` are in `/etc/rc.d`, either because you
! 184: have `PKG_RCD_SCRIPTS=yes`, or manually. (This is not special to Xen,
! 185: but a normal part of pkgsrc usage.)
1.159 maxv 186:
1.186 ! gdt 187: Set `xencommons=YES` in rc.conf:
1.31 gdt 188:
1.157 maxv 189: [[!template id=filecontent name="/etc/rc.conf" text="""
1.152 maxv 190: xencommons=YES
191: """]]
1.31 gdt 192:
1.158 maxv 193: Now, reboot so that you are running a DOM0 kernel under Xen, rather
194: than GENERIC without Xen.
1.157 maxv 195:
1.31 gdt 196: TODO: Recommend for/against xen-watchdog.
1.27 jnemeth 197:
1.158 maxv 198: Once the reboot is done, use `xl` to inspect Xen's boot messages,
1.157 maxv 199: available resources, and running domains. For example:
1.34 gdt 200:
1.153 maxv 201: [[!template id=programlisting text="""
202: # xl dmesg
203: ... xen's boot info ...
204: # xl info
205: ... available memory, etc ...
206: # xl list
207: Name Id Mem(MB) CPU State Time(s) Console
208: Domain-0 0 64 0 r---- 58.1
209: """]]
210:
211: Xen logs will be in /var/log/xen.
1.33 gdt 212:
1.88 gdt 213: ### Issues with xencommons
214:
1.157 maxv 215: `xencommons` starts `xenstored`, which stores data on behalf of dom0 and
1.88 gdt 216: domUs. It does not currently work to stop and start xenstored.
217: Certainly all domUs should be shutdown first, following the sort order
218: of the rc.d scripts. However, the dom0 sets up state with xenstored,
219: and is not notified when xenstored exits, leading to not recreating
220: the state when the new xenstored starts. Until there's a mechanism to
221: make this work, one should not expect to be able to restart xenstored
222: (and thus xencommons). There is currently no reason to expect that
223: this will get fixed any time soon.
224:
1.181 gdt 225: ## anita (for testing NetBSD)
1.41 gdt 226:
1.178 gdt 227: With the setup so far, one should be able to run
1.82 gdt 228: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
229: root, because anita must create a domU):
230:
1.153 maxv 231: [[!template id=programlisting text="""
232: anita --vmm=xl test file:///usr/obj/i386/
233: """]]
1.82 gdt 234:
1.181 gdt 235: ## Xen-specific NetBSD issues
1.40 gdt 236:
237: There are (at least) two additional things different about NetBSD as a
238: dom0 kernel compared to hardware.
239:
1.111 wiz 240: One is that the module ABI is different because some of the #defines
1.109 gdt 241: change, so one must build modules for Xen. As of netbsd-7, the build
1.157 maxv 242: system does this automatically.
1.40 gdt 243:
244: The other difference is that XEN3_DOM0 does not have exactly the same
245: options as GENERIC. While it is debatable whether or not this is a
246: bug, users should be aware of this and can simply add missing config
247: items if desired.
248:
1.181 gdt 249: ## Updating NetBSD in a dom0
1.15 gdt 250:
251: This is just like updating NetBSD on bare hardware, assuming the new
252: version supports the version of Xen you are running. Generally, one
253: replaces the kernel and reboots, and then overlays userland binaries
1.157 maxv 254: and adjusts `/etc`.
1.15 gdt 255:
256: Note that one must update both the non-Xen kernel typically used for
257: rescue purposes and the DOM0 kernel used with Xen.
258:
1.181 gdt 259: ## Converting from grub to /boot
1.55 gdt 260:
1.157 maxv 261: These instructions were used to convert a system from
1.55 gdt 262: grub to /boot. The system was originally installed in February of
263: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
264: over time. Before these commands, it was running NetBSD 6 i386, Xen
265: 4.1 and grub, much like the message linked earlier in the grub
266: section.
267:
1.157 maxv 268: [[!template id=programlisting text="""
269: # Install MBR bootblocks on both disks.
270: fdisk -i /dev/rwd0d
271: fdisk -i /dev/rwd1d
272: # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
273: installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
274: installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
275: # Install secondary boot loader
276: cp -p /usr/mdec/boot /
277: # Create boot.cfg following earlier guidance:
278: menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
279: menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=512M
280: menu=GENERIC:boot
281: menu=GENERIC single-user:boot -s
282: menu=GENERIC.ok:boot netbsd.ok
283: menu=GENERIC.ok single-user:boot netbsd.ok -s
284: menu=Drop to boot prompt:prompt
285: default=1
286: timeout=30
287: """]]
1.22 gdt 288:
1.181 gdt 289: ## Upgrading Xen versions
1.15 gdt 290:
1.110 gdt 291: Minor version upgrades are trivial. Just rebuild/replace the
1.157 maxv 292: xenkernel version and copy the new xen.gz to `/` (where `/boot.cfg`
1.110 gdt 293: references it), and reboot.
294:
1.160 maxv 295: #Unprivileged domains (domU)
1.28 gdt 296:
297: This section describes general concepts about domUs. It does not
1.33 gdt 298: address specific domU operating systems or how to install them. The
1.157 maxv 299: config files for domUs are typically in `/usr/pkg/etc/xen`, and are
1.60 wiki 300: typically named so that the file name, domU name and the domU's host
1.33 gdt 301: name match.
302:
1.111 wiz 303: The domU is provided with CPU and memory by Xen, configured by the
1.33 gdt 304: dom0. The domU is provided with disk and network by the dom0,
305: mediated by Xen, and configured in the dom0.
306:
307: Entropy in domUs can be an issue; physical disks and network are on
308: the dom0. NetBSD's /dev/random system works, but is often challenged.
309:
1.181 gdt 310: ## Config files
1.48 gdt 311:
1.165 gson 312: See /usr/pkg/share/examples/xen/xlexample*
1.164 gson 313: for a small number of well-commented examples, mostly for running
1.48 gdt 314: GNU/Linux.
315:
1.157 maxv 316: The following is an example minimal domain configuration file. The domU
317: serves as a network file server.
318:
319: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
320: name = "domU-id"
321: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
322: memory = 1024
323: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
324: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
325: 'file:/n0/xen/foo-wd1,0x1,w' ]
326: """]]
1.48 gdt 327:
1.157 maxv 328: The domain will have name given in the `name` setting. The kernel has the
1.48 gdt 329: host/domU name in it, so that on the dom0 one can update the various
1.157 maxv 330: domUs independently. The `vif` line causes an interface to be provided,
1.48 gdt 331: with a specific mac address (do not reuse MAC addresses!), in bridge
332: mode. Two disks are provided, and they are both writable; the bits
333: are stored in files and Xen attaches them to a vnd(4) device in the
1.111 wiz 334: dom0 on domain creation. The system treats xbd0 as the boot device
1.48 gdt 335: without needing explicit configuration.
336:
1.163 gson 337: By convention, domain config files are kept in `/usr/pkg/etc/xen`. Note
1.157 maxv 338: that "xl create" takes the name of a config file, while other commands
339: take the name of a domain.
340:
341: Examples of commands:
1.48 gdt 342:
1.157 maxv 343: [[!template id=programlisting text="""
1.163 gson 344: xl create /usr/pkg/etc/xen/foo
345: xl console domU-id
346: xl create -c /usr/pkg/etc/xen/foo
347: xl shutdown domU-id
1.157 maxv 348: xl list
349: """]]
350:
351: Typing `^]` will exit the console session. Shutting down a domain is
1.48 gdt 352: equivalent to pushing the power button; a NetBSD domU will receive a
353: power-press event and do a clean shutdown. Shutting down the dom0
354: will trigger controlled shutdowns of all configured domUs.
355:
1.181 gdt 356: ## CPU and memory
1.33 gdt 357:
1.169 plunky 358: A domain is provided with some number of vcpus, up to the number
1.157 maxv 359: of CPUs seen by the hypervisor. For a domU, it is controlled
1.48 gdt 360: from the config file by the "vcpus = N" directive.
361:
362: A domain is provided with memory; this is controlled in the config
363: file by "memory = N" (in megabytes). In the straightforward case, the
364: sum of the the memory allocated to the dom0 and all domUs must be less
1.33 gdt 365: than the available memory.
366:
367: Xen also provides a "balloon" driver, which can be used to let domains
1.157 maxv 368: use more memory temporarily.
1.28 gdt 369:
1.181 gdt 370: ## Virtual disks
1.28 gdt 371:
1.158 maxv 372: In domU config files, the disks are defined as a sequence of 3-tuples:
373:
374: * The first element is "method:/path/to/disk". Common methods are
1.162 gutterid 375: "file:" for a file-backed vnd, and "phy:" for something that is already
376: a device, such as an LVM logical volume.
1.158 maxv 377:
378: * The second element is an artifact of how virtual disks are passed to
379: Linux, and a source of confusion with NetBSD Xen usage. Linux domUs
380: are given a device name to associate with the disk, and values like
381: "hda1" or "sda1" are common. In a NetBSD domU, the first disk appears
382: as xbd0, the second as xbd1, and so on. However, xl demands a
383: second argument. The name given is converted to a major/minor by
384: calling stat(2) on the name in /dev and this is passed to the domU.
385: In the general case, the dom0 and domU can be different operating
386: systems, and it is an unwarranted assumption that they have consistent
387: numbering in /dev, or even that the dom0 OS has a /dev. With NetBSD
388: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
389: for the second works fine and avoids this issue. For a GNU/Linux
390: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
391: /dev/hda1.
1.48 gdt 392:
1.158 maxv 393: * The third element is "w" for writable disks, and "r" for read-only
394: disks.
395:
396: Example:
397: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
398: disk = [ 'file:/n0/xen/foo-wd0,0x0,w' ]
399: """]]
1.28 gdt 400:
1.127 gdt 401: Note that NetBSD by default creates only vnd[0123]. If you need more
402: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
403: dom0.
404:
405: Note that NetBSD by default creates only xbd[0123]. If you need more
406: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
407:
1.28 gdt 408: Virtual Networking
409: ------------------
410:
1.111 wiz 411: Xen provides virtual Ethernets, each of which connects the dom0 and a
1.46 gdt 412: domU. For each virtual network, there is an interface "xvifN.M" in
1.158 maxv 413: the dom0, and a matching interface xennetM (NetBSD name) in domU index N.
414: The interfaces behave as if there is an Ethernet with two
1.111 wiz 415: adapters connected. From this primitive, one can construct various
1.46 gdt 416: configurations. We focus on two common and useful cases for which
417: there are existing scripts: bridging and NAT.
1.28 gdt 418:
1.48 gdt 419: With bridging (in the example above), the domU perceives itself to be
420: on the same network as the dom0. For server virtualization, this is
421: usually best. Bridging is accomplished by creating a bridge(4) device
422: and adding the dom0's physical interface and the various xvifN.0
423: interfaces to the bridge. One specifies "bridge=bridge0" in the domU
424: config file. The bridge must be set up already in the dom0; an
425: example /etc/ifconfig.bridge0 is:
1.46 gdt 426:
1.157 maxv 427: [[!template id=filecontent name="/etc/ifconfig.bridge0" text="""
428: create
429: up
430: !brconfig bridge0 add wm0
431: """]]
1.28 gdt 432:
433: With NAT, the domU perceives itself to be behind a NAT running on the
434: dom0. This is often appropriate when running Xen on a workstation.
1.48 gdt 435: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28 gdt 436:
1.49 gdt 437: The MAC address specified is the one used for the interface in the new
1.53 gdt 438: domain. The interface in dom0 will use this address XOR'd with
1.49 gdt 439: 00:00:00:01:00:00. Random MAC addresses are assigned if not given.
440:
1.48 gdt 441: Starting domains automatically
442: ------------------------------
1.28 gdt 443:
1.158 maxv 444: To start domains `domU-netbsd` and `domU-linux` at boot and shut them
445: down cleanly on dom0 shutdown, add the following in rc.conf:
1.28 gdt 446:
1.158 maxv 447: [[!template id=filecontent name="/etc/rc.conf" text="""
448: xendomains="domU-netbsd domU-linux"
449: """]]
1.28 gdt 450:
1.180 gdt 451: # Creating a domU
1.14 gdt 452:
453: Creating domUs is almost entirely independent of operating system. We
1.49 gdt 454: have already presented the basics of config files. Note that you must
1.158 maxv 455: have already completed the dom0 setup so that "xl list" works.
1.14 gdt 456:
1.180 gdt 457: Creating a NetBSD PV domU
458: --------------------------
1.1 mspo 459:
1.49 gdt 460: See the earlier config file, and adjust memory. Decide on how much
1.162 gutterid 461: storage you will provide, and prepare it (file or LVM).
1.49 gdt 462:
1.111 wiz 463: While the kernel will be obtained from the dom0 file system, the same
1.49 gdt 464: file should be present in the domU as /netbsd so that tools like
465: savecore(8) can work. (This is helpful but not necessary.)
466:
467: The kernel must be specifically for Xen and for use as a domU. The
468: i386 and amd64 provide the following kernels:
469:
470: i386 XEN3PAE_DOMU
1.95 gdt 471: amd64 XEN3_DOMU
1.5 mspo 472:
1.49 gdt 473: This will boot NetBSD, but this is not that useful if the disk is
474: empty. One approach is to unpack sets onto the disk outside of xen
475: (by mounting it, just as you would prepare a physical disk for a
476: system you can't run the installer on).
477:
478: A second approach is to run an INSTALL kernel, which has a miniroot
479: and can load sets from the network. To do this, copy the INSTALL
480: kernel to / and change the kernel line in the config file to:
1.5 mspo 481:
1.49 gdt 482: kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5 mspo 483:
1.163 gson 484: Then, start the domain as "xl create -c configfile".
1.1 mspo 485:
1.49 gdt 486: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
487: line should be used in the config file.
1.1 mspo 488:
1.3 mspo 489: disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1 mspo 490:
491: After booting the domain, the option to install via CDROM may be
1.49 gdt 492: selected. The CDROM device should be changed to `xbd1d`.
1.1 mspo 493:
1.49 gdt 494: Once done installing, "halt -p" the new domain (don't reboot or halt,
495: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
496: config file), switch the config file back to the XEN3_DOMU kernel,
497: and start the new domain again. Now it should be able to use "root on
498: xbd0a" and you should have a, functional NetBSD domU.
1.1 mspo 499:
1.49 gdt 500: TODO: check if this is still accurate.
1.1 mspo 501: When the new domain is booting you'll see some warnings about *wscons*
502: and the pseudo-terminals. These can be fixed by editing the files
1.5 mspo 503: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
504: `/etc/ttys`, except *console*, like this:
1.1 mspo 505:
1.3 mspo 506: console "/usr/libexec/getty Pc" vt100 on secure
507: ttyE0 "/usr/libexec/getty Pc" vt220 off secure
508: ttyE1 "/usr/libexec/getty Pc" vt220 off secure
509: ttyE2 "/usr/libexec/getty Pc" vt220 off secure
510: ttyE3 "/usr/libexec/getty Pc" vt220 off secure
1.1 mspo 511:
1.5 mspo 512: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1 mspo 513:
514: It is also desirable to add
515:
1.49 gdt 516: powerd=YES
1.1 mspo 517:
1.5 mspo 518: in rc.conf. This way, the domain will be properly shut down if
1.184 gdt 519: `xl shutdown -R` or `xl shutdown -H` is used on the dom0.
520: \todo Check the translation to xl.
1.1 mspo 521:
1.92 gdt 522: It is not strictly necessary to have a kernel (as /netbsd) in the domU
1.111 wiz 523: file system. However, various programs (e.g. netstat) will use that
1.92 gdt 524: kernel to look up symbols to read from kernel virtual memory. If
525: /netbsd is not the running kernel, those lookups will fail. (This is
526: not really a Xen-specific issue, but because the domU kernel is
527: obtained from the dom0, it is far more likely to be out of sync or
528: missing with Xen.)
529:
1.158 maxv 530: Creating a Linux domU
531: ---------------------
1.1 mspo 532:
533: Creating unprivileged Linux domains isn't much different from
534: unprivileged NetBSD domains, but there are some details to know.
535:
536: First, the second parameter passed to the disk declaration (the '0x1' in
537: the example below)
538:
1.3 mspo 539: disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1 mspo 540:
541: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49 gdt 542: for hda). Linux builds device numbers as: (major \<\< 8 + minor).
543: So, hda1 which has major 3 and minor 1 on a Linux system will have
544: device number 0x301. Alternatively, devices names can be used (hda,
545: hdb, ...) as xentools has a table to map these names to devices
546: numbers. To export a partition to a Linux guest we can use:
1.1 mspo 547:
1.49 gdt 548: disk = [ 'phy:/dev/wd0e,0x300,w' ]
549: root = "/dev/hda1 ro"
1.1 mspo 550:
551: and it will appear as /dev/hda on the Linux system, and be used as root
552: partition.
553:
1.49 gdt 554: To install the Linux system on the partition to be exported to the
555: guest domain, the following method can be used: install
556: sysutils/e2fsprogs from pkgsrc. Use mke2fs to format the partition
557: that will be the root partition of your Linux domain, and mount it.
558: Then copy the files from a working Linux system, make adjustments in
559: `/etc` (fstab, network config). It should also be possible to extract
560: binary packages such as .rpm or .deb directly to the mounted partition
561: using the appropriate tool, possibly running under NetBSD's Linux
1.111 wiz 562: emulation. Once the file system has been populated, umount it. If
563: desirable, the file system can be converted to ext3 using tune2fs -j.
1.49 gdt 564: It should now be possible to boot the Linux guest domain, using one of
565: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1 mspo 566:
1.111 wiz 567: To get the Linux console right, you need to add:
1.1 mspo 568:
1.3 mspo 569: extra = "xencons=tty1"
1.1 mspo 570:
1.111 wiz 571: to your configuration since not all Linux distributions auto-attach a
1.1 mspo 572: tty to the xen console.
573:
1.180 gdt 574: ## Creating a NetBSD HVM domU
575:
576: Use type='hmv', probably. Use a GENERIC kernel within the disk image.
577:
578: ## Creating a NetBSD PVH domU
579:
580: Use type='pvh'.
581:
582: \todo Explain where the kernel comes from.
583:
584:
1.158 maxv 585: Creating a Solaris domU
586: -----------------------
1.1 mspo 587:
1.50 gdt 588: See possibly outdated
589: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5 mspo 590:
1.1 mspo 591:
1.52 gdt 592: PCI passthrough: Using PCI devices in guest domains
593: ---------------------------------------------------
1.1 mspo 594:
1.180 gdt 595: NB: PCI passthrough only works on some Xen versions and as of 2020 it
596: is not clear that it works on any version in pkgsrc. Reports
597: confirming or denying this notion should be sent to port-xen@.
598:
1.53 gdt 599: The dom0 can give other domains access to selected PCI
1.52 gdt 600: devices. This can allow, for example, a non-privileged domain to have
601: access to a physical network interface or disk controller. However,
602: keep in mind that giving a domain access to a PCI device most likely
603: will give the domain read/write access to the whole physical memory,
604: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53 gdt 605: device. Also, it's not possible to export ISA devices to non-dom0
1.52 gdt 606: domains, which means that the primary VGA adapter can't be exported.
607: A guest domain trying to access the VGA registers will panic.
608:
1.53 gdt 609: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52 gdt 610: not been ported to later versions at this time.
611:
612: For a PCI device to be exported to a domU, is has to be attached to
613: the "pciback" driver in dom0. Devices passed to the dom0 via the
614: pciback.hide boot parameter will attach to "pciback" instead of the
615: usual driver. The list of devices is specified as "(bus:dev.func)",
1.5 mspo 616: where bus and dev are 2-digit hexadecimal numbers, and func a
617: single-digit number:
1.1 mspo 618:
1.52 gdt 619: pciback.hide=(00:0a.0)(00:06.0)
1.1 mspo 620:
1.52 gdt 621: pciback devices should show up in the dom0's boot messages, and the
1.5 mspo 622: devices should be listed in the `/kern/xen/pci` directory.
1.1 mspo 623:
1.52 gdt 624: PCI devices to be exported to a domU are listed in the "pci" array of
625: the domU's config file, with the format "0000:bus:dev.func".
1.1 mspo 626:
1.52 gdt 627: pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1 mspo 628:
1.52 gdt 629: In the domU an "xpci" device will show up, to which one or more pci
1.111 wiz 630: buses will attach. Then the PCI drivers will attach to PCI buses as
1.52 gdt 631: usual. Note that the default NetBSD DOMU kernels do not have "xpci"
632: or any PCI drivers built in by default; you have to build your own
633: kernel to use PCI devices in a domU. Here's a kernel config example;
634: note that only the "xpci" lines are unusual.
635:
636: include "arch/i386/conf/XEN3_DOMU"
637:
1.111 wiz 638: # Add support for PCI buses to the XEN3_DOMU kernel
1.52 gdt 639: xpci* at xenbus ?
640: pci* at xpci ?
641:
642: # PCI USB controllers
643: uhci* at pci? dev ? function ? # Universal Host Controller (Intel)
644:
645: # USB bus support
646: usb* at uhci?
647:
648: # USB Hubs
649: uhub* at usb?
650: uhub* at uhub? port ? configuration ? interface ?
651:
652: # USB Mass Storage
653: umass* at uhub? port ? configuration ? interface ?
654: wd* at umass?
655: # SCSI controllers
656: ahc* at pci? dev ? function ? # Adaptec [23]94x, aic78x0 SCSI
657:
658: # SCSI bus support (for both ahc and umass)
659: scsibus* at scsi?
660:
661: # SCSI devices
662: sd* at scsibus? target ? lun ? # SCSI disk drives
663: cd* at scsibus? target ? lun ? # SCSI CD-ROM drives
1.1 mspo 664:
665:
1.173 gdt 666: # Specific Issues
667:
668: ## domU
669:
1.176 gdt 670: [NetBSD 5 is known to panic.](http://mail-index.netbsd.org/port-xen/2018/04/17/msg009181.html)
1.175 gdt 671: (However, NetBSD 5 systems should be updated to a supported version.)
1.173 gdt 672:
673: # NetBSD as a domU in a VPS
1.28 gdt 674:
675: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
676: hardware. This section explains how to deal with Xen in a domU as a
677: virtual private server where you do not control or have access to the
1.70 gdt 678: dom0. This is not intended to be an exhaustive list of VPS providers;
679: only a few are mentioned that specifically support NetBSD.
1.28 gdt 680:
1.52 gdt 681: VPS operators provide varying degrees of access and mechanisms for
682: configuration. The big issue is usually how one controls which kernel
1.111 wiz 683: is booted, because the kernel is nominally in the dom0 file system (to
684: which VPS users do not normally have access). A second issue is how
1.70 gdt 685: to install NetBSD.
1.52 gdt 686: A VPS user may want to compile a kernel for security updates, to run
687: npf, run IPsec, or any other reason why someone would want to change
688: their kernel.
689:
1.111 wiz 690: One approach is to have an administrative interface to upload a kernel,
1.68 gdt 691: or to select from a prepopulated list. Other approaches are pygrub
1.59 gdt 692: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
1.111 wiz 693: kernel from the domU file system. This is closer to a regular physical
1.59 gdt 694: computer, where someone who controls a machine can replace the kernel.
1.52 gdt 695:
1.74 gdt 696: A second issue is multiple CPUs. With NetBSD 6, domUs support
697: multiple vcpus, and it is typical for VPS providers to enable multiple
698: CPUs for NetBSD domUs.
699:
1.180 gdt 700: ## Complexities due to Xen changes
701:
702: Xen has many security advisories and people running Xen systems make
703: different choices.
704:
705: ### stub domains
706:
707: Some (Linux only?) dom0 systems use something called "stub domains" to
708: isolate qemu from the dom0 system, as a security and reliabilty
709: mechanism when running HVM domUs. Somehow, NetBSD's GENERIC kernel
710: ends up using PIO for disks rather than DMA. Of course, all of this
711: is emulated, but emulated PIO is unusably slow. This problem is not
712: currently understood.
713:
714: ### Grant tables
715:
716: There are multiple versions of using grant tables, and some security
717: advisories have suggested disabling some versions. Some versions of
718: NetBSD apparently only use specific versions and this can lead to
719: "NetBSD current doesn't run on hosting provider X" situations.
720:
721: \todo Explain better.
722:
1.181 gdt 723: ## Boot methods
724:
725: ### pvgrub
1.59 gdt 726:
727: pvgrub is a version of grub that uses PV operations instead of BIOS
728: calls. It is booted from the dom0 as the domU kernel, and then reads
1.111 wiz 729: /grub/menu.lst and loads a kernel from the domU file system.
1.59 gdt 730:
1.70 gdt 731: [Panix](http://www.panix.com/) lets users use pvgrub. Panix reports
1.71 gdt 732: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
733: (and hence with defaults from "newfs -O 2"). See [Panix's pvgrub
1.70 gdt 734: page](http://www.panix.com/v-colo/grub.html), which describes only
1.74 gdt 735: Linux but should be updated to cover NetBSD :-).
1.70 gdt 736:
737: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
738: their own kernel. See then [prgmr.com NetBSD
1.74 gdt 739: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
740: (which is in need of updating).
1.59 gdt 741:
1.70 gdt 742: It appears that [grub's FFS
743: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
744: does not support all aspects of modern FFS, but there are also reports
1.72 gdt 745: that FFSv2 works fine. At prgmr, typically one has an ext2 or FAT
1.70 gdt 746: partition for the kernel with the intent that grub can understand it,
747: which leads to /netbsd not being the actual kernel. One must remember
1.111 wiz 748: to update the special boot partition.
1.59 gdt 749:
1.181 gdt 750: ### pygrub
1.168 wiki 751:
752: pygrub runs in the dom0 and looks into the domU file system. This
753: implies that the domU must have a kernel in a file system in a format
754: known to pygrub.
755:
756: pygrub doesn't seem to work to load Linux images under NetBSD dom0,
757: and is inherently less secure than pvgrub due to running inside dom0. For both these
758: reasons, pygrub should not be used, and is only still present so that
759: historical DomU images using it still work.
760:
761: As of 2014, pygrub seems to be of mostly historical
762: interest. New DomUs should use pvgrub.
763:
1.181 gdt 764: ## Specific Providers
765:
766: ### Amazon
1.59 gdt 767:
1.143 wiki 768: See the [Amazon EC2 page](/amazon_ec2/).
CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb