wikisrc/ports/xen/howto.mdwn - view

File: [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.82: download - view: text, annotated - select for diffs
Sat Jan 17 13:11:49 2015 UTC (6 years ago) by gdt
Branches: MAIN
CVS tags: HEAD

add xen/qemu example

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