You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

qm.adoc 53KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347
  1. [[chapter_virtual_machines]]
  2. ifdef::manvolnum[]
  3. qm(1)
  4. =====
  5. :pve-toplevel:
  6. NAME
  7. ----
  8. qm - Qemu/KVM Virtual Machine Manager
  9. SYNOPSIS
  10. --------
  11. include::qm.1-synopsis.adoc[]
  12. DESCRIPTION
  13. -----------
  14. endif::manvolnum[]
  15. ifndef::manvolnum[]
  16. Qemu/KVM Virtual Machines
  17. =========================
  18. :pve-toplevel:
  19. endif::manvolnum[]
  20. // deprecates
  21. // http://pve.proxmox.com/wiki/Container_and_Full_Virtualization
  22. // http://pve.proxmox.com/wiki/KVM
  23. // http://pve.proxmox.com/wiki/Qemu_Server
  24. Qemu (short form for Quick Emulator) is an open source hypervisor that emulates a
  25. physical computer. From the perspective of the host system where Qemu is
  26. running, Qemu is a user program which has access to a number of local resources
  27. like partitions, files, network cards which are then passed to an
  28. emulated computer which sees them as if they were real devices.
  29. A guest operating system running in the emulated computer accesses these
  30. devices, and runs as it were running on real hardware. For instance you can pass
  31. an iso image as a parameter to Qemu, and the OS running in the emulated computer
  32. will see a real CDROM inserted in a CD drive.
  33. Qemu can emulate a great variety of hardware from ARM to Sparc, but {pve} is
  34. only concerned with 32 and 64 bits PC clone emulation, since it represents the
  35. overwhelming majority of server hardware. The emulation of PC clones is also one
  36. of the fastest due to the availability of processor extensions which greatly
  37. speed up Qemu when the emulated architecture is the same as the host
  38. architecture.
  39. NOTE: You may sometimes encounter the term _KVM_ (Kernel-based Virtual Machine).
  40. It means that Qemu is running with the support of the virtualization processor
  41. extensions, via the Linux kvm module. In the context of {pve} _Qemu_ and
  42. _KVM_ can be used interchangeably as Qemu in {pve} will always try to load the kvm
  43. module.
  44. Qemu inside {pve} runs as a root process, since this is required to access block
  45. and PCI devices.
  46. Emulated devices and paravirtualized devices
  47. --------------------------------------------
  48. The PC hardware emulated by Qemu includes a mainboard, network controllers,
  49. scsi, ide and sata controllers, serial ports (the complete list can be seen in
  50. the `kvm(1)` man page) all of them emulated in software. All these devices
  51. are the exact software equivalent of existing hardware devices, and if the OS
  52. running in the guest has the proper drivers it will use the devices as if it
  53. were running on real hardware. This allows Qemu to runs _unmodified_ operating
  54. systems.
  55. This however has a performance cost, as running in software what was meant to
  56. run in hardware involves a lot of extra work for the host CPU. To mitigate this,
  57. Qemu can present to the guest operating system _paravirtualized devices_, where
  58. the guest OS recognizes it is running inside Qemu and cooperates with the
  59. hypervisor.
  60. Qemu relies on the virtio virtualization standard, and is thus able to present
  61. paravirtualized virtio devices, which includes a paravirtualized generic disk
  62. controller, a paravirtualized network card, a paravirtualized serial port,
  63. a paravirtualized SCSI controller, etc ...
  64. It is highly recommended to use the virtio devices whenever you can, as they
  65. provide a big performance improvement. Using the virtio generic disk controller
  66. versus an emulated IDE controller will double the sequential write throughput,
  67. as measured with `bonnie++(8)`. Using the virtio network interface can deliver
  68. up to three times the throughput of an emulated Intel E1000 network card, as
  69. measured with `iperf(1)`. footnote:[See this benchmark on the KVM wiki
  70. http://www.linux-kvm.org/page/Using_VirtIO_NIC]
  71. [[qm_virtual_machines_settings]]
  72. Virtual Machines Settings
  73. -------------------------
  74. Generally speaking {pve} tries to choose sane defaults for virtual machines
  75. (VM). Make sure you understand the meaning of the settings you change, as it
  76. could incur a performance slowdown, or putting your data at risk.
  77. [[qm_general_settings]]
  78. General Settings
  79. ~~~~~~~~~~~~~~~~
  80. [thumbnail="screenshot/gui-create-vm-general.png"]
  81. General settings of a VM include
  82. * the *Node* : the physical server on which the VM will run
  83. * the *VM ID*: a unique number in this {pve} installation used to identify your VM
  84. * *Name*: a free form text string you can use to describe the VM
  85. * *Resource Pool*: a logical group of VMs
  86. [[qm_os_settings]]
  87. OS Settings
  88. ~~~~~~~~~~~
  89. [thumbnail="screenshot/gui-create-vm-os.png"]
  90. When creating a virtual machine (VM), setting the proper Operating System(OS)
  91. allows {pve} to optimize some low level parameters. For instance Windows OS
  92. expect the BIOS clock to use the local time, while Unix based OS expect the
  93. BIOS clock to have the UTC time.
  94. [[qm_system_settings]]
  95. System Settings
  96. ~~~~~~~~~~~~~~~
  97. On VM creation you can change some basic system components of the new VM. You
  98. can specify which xref:qm_display[display type] you want to use.
  99. [thumbnail="screenshot/gui-create-vm-system.png"]
  100. Additionally, the xref:qm_hard_disk[SCSI controller] can be changed.
  101. If you plan to install the QEMU Guest Agent, or if your selected ISO image
  102. already ships and installs it automatically, you may want to tick the 'Qemu
  103. Agent' box, which lets {pve} know that it can use its features to show some
  104. more information, and complete some actions (for example, shutdown or
  105. snapshots) more intelligently.
  106. {pve} allows to boot VMs with different firmware and machine types, namely
  107. xref:qm_bios_and_uefi[SeaBIOS and OVMF]. In most cases you want to switch from
  108. the default SeabBIOS to OVMF only if you plan to use
  109. xref:qm_pci_passthrough[PCIe pass through]. A VMs 'Machine Type' defines the
  110. hardware layout of the VM's virtual motherboard. You can choose between the
  111. default https://en.wikipedia.org/wiki/Intel_440FX[Intel 440FX] or the
  112. https://ark.intel.com/content/www/us/en/ark/products/31918/intel-82q35-graphics-and-memory-controller.html[Q35]
  113. chipset, which also provides a virtual PCIe bus, and thus may be desired if
  114. one want's to pass through PCIe hardware.
  115. [[qm_hard_disk]]
  116. Hard Disk
  117. ~~~~~~~~~
  118. [[qm_hard_disk_bus]]
  119. Bus/Controller
  120. ^^^^^^^^^^^^^^
  121. Qemu can emulate a number of storage controllers:
  122. * the *IDE* controller, has a design which goes back to the 1984 PC/AT disk
  123. controller. Even if this controller has been superseded by recent designs,
  124. each and every OS you can think of has support for it, making it a great choice
  125. if you want to run an OS released before 2003. You can connect up to 4 devices
  126. on this controller.
  127. * the *SATA* (Serial ATA) controller, dating from 2003, has a more modern
  128. design, allowing higher throughput and a greater number of devices to be
  129. connected. You can connect up to 6 devices on this controller.
  130. * the *SCSI* controller, designed in 1985, is commonly found on server grade
  131. hardware, and can connect up to 14 storage devices. {pve} emulates by default a
  132. LSI 53C895A controller.
  133. +
  134. A SCSI controller of type _VirtIO SCSI_ is the recommended setting if you aim for
  135. performance and is automatically selected for newly created Linux VMs since
  136. {pve} 4.3. Linux distributions have support for this controller since 2012, and
  137. FreeBSD since 2014. For Windows OSes, you need to provide an extra iso
  138. containing the drivers during the installation.
  139. // https://pve.proxmox.com/wiki/Paravirtualized_Block_Drivers_for_Windows#During_windows_installation.
  140. If you aim at maximum performance, you can select a SCSI controller of type
  141. _VirtIO SCSI single_ which will allow you to select the *IO Thread* option.
  142. When selecting _VirtIO SCSI single_ Qemu will create a new controller for
  143. each disk, instead of adding all disks to the same controller.
  144. * The *VirtIO Block* controller, often just called VirtIO or virtio-blk,
  145. is an older type of paravirtualized controller. It has been superseded by the
  146. VirtIO SCSI Controller, in terms of features.
  147. [thumbnail="screenshot/gui-create-vm-hard-disk.png"]
  148. [[qm_hard_disk_formats]]
  149. Image Format
  150. ^^^^^^^^^^^^
  151. On each controller you attach a number of emulated hard disks, which are backed
  152. by a file or a block device residing in the configured storage. The choice of
  153. a storage type will determine the format of the hard disk image. Storages which
  154. present block devices (LVM, ZFS, Ceph) will require the *raw disk image format*,
  155. whereas files based storages (Ext4, NFS, CIFS, GlusterFS) will let you to choose
  156. either the *raw disk image format* or the *QEMU image format*.
  157. * the *QEMU image format* is a copy on write format which allows snapshots, and
  158. thin provisioning of the disk image.
  159. * the *raw disk image* is a bit-to-bit image of a hard disk, similar to what
  160. you would get when executing the `dd` command on a block device in Linux. This
  161. format does not support thin provisioning or snapshots by itself, requiring
  162. cooperation from the storage layer for these tasks. It may, however, be up to
  163. 10% faster than the *QEMU image format*. footnote:[See this benchmark for details
  164. http://events.linuxfoundation.org/sites/events/files/slides/CloudOpen2013_Khoa_Huynh_v3.pdf]
  165. * the *VMware image format* only makes sense if you intend to import/export the
  166. disk image to other hypervisors.
  167. [[qm_hard_disk_cache]]
  168. Cache Mode
  169. ^^^^^^^^^^
  170. Setting the *Cache* mode of the hard drive will impact how the host system will
  171. notify the guest systems of block write completions. The *No cache* default
  172. means that the guest system will be notified that a write is complete when each
  173. block reaches the physical storage write queue, ignoring the host page cache.
  174. This provides a good balance between safety and speed.
  175. If you want the {pve} backup manager to skip a disk when doing a backup of a VM,
  176. you can set the *No backup* option on that disk.
  177. If you want the {pve} storage replication mechanism to skip a disk when starting
  178. a replication job, you can set the *Skip replication* option on that disk.
  179. As of {pve} 5.0, replication requires the disk images to be on a storage of type
  180. `zfspool`, so adding a disk image to other storages when the VM has replication
  181. configured requires to skip replication for this disk image.
  182. [[qm_hard_disk_discard]]
  183. Trim/Discard
  184. ^^^^^^^^^^^^
  185. If your storage supports _thin provisioning_ (see the storage chapter in the
  186. {pve} guide), you can activate the *Discard* option on a drive. With *Discard*
  187. set and a _TRIM_-enabled guest OS footnote:[TRIM, UNMAP, and discard
  188. https://en.wikipedia.org/wiki/Trim_%28computing%29], when the VM's filesystem
  189. marks blocks as unused after deleting files, the controller will relay this
  190. information to the storage, which will then shrink the disk image accordingly.
  191. For the guest to be able to issue _TRIM_ commands, you must enable the *Discard*
  192. option on the drive. Some guest operating systems may also require the
  193. *SSD Emulation* flag to be set. Note that *Discard* on *VirtIO Block* drives is
  194. only supported on guests using Linux Kernel 5.0 or higher.
  195. If you would like a drive to be presented to the guest as a solid-state drive
  196. rather than a rotational hard disk, you can set the *SSD emulation* option on
  197. that drive. There is no requirement that the underlying storage actually be
  198. backed by SSDs; this feature can be used with physical media of any type.
  199. Note that *SSD emulation* is not supported on *VirtIO Block* drives.
  200. [[qm_hard_disk_iothread]]
  201. IO Thread
  202. ^^^^^^^^^
  203. The option *IO Thread* can only be used when using a disk with the
  204. *VirtIO* controller, or with the *SCSI* controller, when the emulated controller
  205. type is *VirtIO SCSI single*.
  206. With this enabled, Qemu creates one I/O thread per storage controller,
  207. instead of a single thread for all I/O, so it can increase performance when
  208. multiple isks are used and each disk has its own storage controller.
  209. [[qm_cpu]]
  210. CPU
  211. ~~~
  212. [thumbnail="screenshot/gui-create-vm-cpu.png"]
  213. A *CPU socket* is a physical slot on a PC motherboard where you can plug a CPU.
  214. This CPU can then contain one or many *cores*, which are independent
  215. processing units. Whether you have a single CPU socket with 4 cores, or two CPU
  216. sockets with two cores is mostly irrelevant from a performance point of view.
  217. However some software licenses depend on the number of sockets a machine has,
  218. in that case it makes sense to set the number of sockets to what the license
  219. allows you.
  220. Increasing the number of virtual cpus (cores and sockets) will usually provide a
  221. performance improvement though that is heavily dependent on the use of the VM.
  222. Multithreaded applications will of course benefit from a large number of
  223. virtual cpus, as for each virtual cpu you add, Qemu will create a new thread of
  224. execution on the host system. If you're not sure about the workload of your VM,
  225. it is usually a safe bet to set the number of *Total cores* to 2.
  226. NOTE: It is perfectly safe if the _overall_ number of cores of all your VMs
  227. is greater than the number of cores on the server (e.g., 4 VMs with each 4
  228. cores on a machine with only 8 cores). In that case the host system will
  229. balance the Qemu execution threads between your server cores, just like if you
  230. were running a standard multithreaded application. However, {pve} will prevent
  231. you from assigning more virtual CPU cores than physically available, as this will
  232. only bring the performance down due to the cost of context switches.
  233. [[qm_cpu_resource_limits]]
  234. Resource Limits
  235. ^^^^^^^^^^^^^^^
  236. In addition to the number of virtual cores, you can configure how much resources
  237. a VM can get in relation to the host CPU time and also in relation to other
  238. VMs.
  239. With the *cpulimit* (``Host CPU Time'') option you can limit how much CPU time
  240. the whole VM can use on the host. It is a floating point value representing CPU
  241. time in percent, so `1.0` is equal to `100%`, `2.5` to `250%` and so on. If a
  242. single process would fully use one single core it would have `100%` CPU Time
  243. usage. If a VM with four cores utilizes all its cores fully it would
  244. theoretically use `400%`. In reality the usage may be even a bit higher as Qemu
  245. can have additional threads for VM peripherals besides the vCPU core ones.
  246. This setting can be useful if a VM should have multiple vCPUs, as it runs a few
  247. processes in parallel, but the VM as a whole should not be able to run all
  248. vCPUs at 100% at the same time. Using a specific example: lets say we have a VM
  249. which would profit from having 8 vCPUs, but at no time all of those 8 cores
  250. should run at full load - as this would make the server so overloaded that
  251. other VMs and CTs would get to less CPU. So, we set the *cpulimit* limit to
  252. `4.0` (=400%). If all cores do the same heavy work they would all get 50% of a
  253. real host cores CPU time. But, if only 4 would do work they could still get
  254. almost 100% of a real core each.
  255. NOTE: VMs can, depending on their configuration, use additional threads e.g.,
  256. for networking or IO operations but also live migration. Thus a VM can show up
  257. to use more CPU time than just its virtual CPUs could use. To ensure that a VM
  258. never uses more CPU time than virtual CPUs assigned set the *cpulimit* setting
  259. to the same value as the total core count.
  260. The second CPU resource limiting setting, *cpuunits* (nowadays often called CPU
  261. shares or CPU weight), controls how much CPU time a VM gets in regards to other
  262. VMs running. It is a relative weight which defaults to `1024`, if you increase
  263. this for a VM it will be prioritized by the scheduler in comparison to other
  264. VMs with lower weight. E.g., if VM 100 has set the default 1024 and VM 200 was
  265. changed to `2048`, the latter VM 200 would receive twice the CPU bandwidth than
  266. the first VM 100.
  267. For more information see `man systemd.resource-control`, here `CPUQuota`
  268. corresponds to `cpulimit` and `CPUShares` corresponds to our `cpuunits`
  269. setting, visit its Notes section for references and implementation details.
  270. CPU Type
  271. ^^^^^^^^
  272. Qemu can emulate a number different of *CPU types* from 486 to the latest Xeon
  273. processors. Each new processor generation adds new features, like hardware
  274. assisted 3d rendering, random number generation, memory protection, etc ...
  275. Usually you should select for your VM a processor type which closely matches the
  276. CPU of the host system, as it means that the host CPU features (also called _CPU
  277. flags_ ) will be available in your VMs. If you want an exact match, you can set
  278. the CPU type to *host* in which case the VM will have exactly the same CPU flags
  279. as your host system.
  280. This has a downside though. If you want to do a live migration of VMs between
  281. different hosts, your VM might end up on a new system with a different CPU type.
  282. If the CPU flags passed to the guest are missing, the qemu process will stop. To
  283. remedy this Qemu has also its own CPU type *kvm64*, that {pve} uses by defaults.
  284. kvm64 is a Pentium 4 look a like CPU type, which has a reduced CPU flags set,
  285. but is guaranteed to work everywhere.
  286. In short, if you care about live migration and moving VMs between nodes, leave
  287. the kvm64 default. If you don’t care about live migration or have a homogeneous
  288. cluster where all nodes have the same CPU, set the CPU type to host, as in
  289. theory this will give your guests maximum performance.
  290. Meltdown / Spectre related CPU flags
  291. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  292. There are several CPU flags related to the Meltdown and Spectre vulnerabilities
  293. footnote:[Meltdown Attack https://meltdownattack.com/] which need to be set
  294. manually unless the selected CPU type of your VM already enables them by default.
  295. There are two requirements that need to be fulfilled in order to use these
  296. CPU flags:
  297. * The host CPU(s) must support the feature and propagate it to the guest's virtual CPU(s)
  298. * The guest operating system must be updated to a version which mitigates the
  299. attacks and is able to utilize the CPU feature
  300. Otherwise you need to set the desired CPU flag of the virtual CPU, either by
  301. editing the CPU options in the WebUI, or by setting the 'flags' property of the
  302. 'cpu' option in the VM configuration file.
  303. For Spectre v1,v2,v4 fixes, your CPU or system vendor also needs to provide a
  304. so-called ``microcode update'' footnote:[You can use `intel-microcode' /
  305. `amd-microcode' from Debian non-free if your vendor does not provide such an
  306. update. Note that not all affected CPUs can be updated to support spec-ctrl.]
  307. for your CPU.
  308. To check if the {pve} host is vulnerable, execute the following command as root:
  309. ----
  310. for f in /sys/devices/system/cpu/vulnerabilities/*; do echo "${f##*/} -" $(cat "$f"); done
  311. ----
  312. A community script is also available to detect is the host is still vulnerable.
  313. footnote:[spectre-meltdown-checker https://meltdown.ovh/]
  314. Intel processors
  315. ^^^^^^^^^^^^^^^^
  316. * 'pcid'
  317. +
  318. This reduces the performance impact of the Meltdown (CVE-2017-5754) mitigation
  319. called 'Kernel Page-Table Isolation (KPTI)', which effectively hides
  320. the Kernel memory from the user space. Without PCID, KPTI is quite an expensive
  321. mechanism footnote:[PCID is now a critical performance/security feature on x86
  322. https://groups.google.com/forum/m/#!topic/mechanical-sympathy/L9mHTbeQLNU].
  323. +
  324. To check if the {pve} host supports PCID, execute the following command as root:
  325. +
  326. ----
  327. # grep ' pcid ' /proc/cpuinfo
  328. ----
  329. +
  330. If this does not return empty your host's CPU has support for 'pcid'.
  331. * 'spec-ctrl'
  332. +
  333. Required to enable the Spectre v1 (CVE-2017-5753) and Spectre v2 (CVE-2017-5715) fix,
  334. in cases where retpolines are not sufficient.
  335. Included by default in Intel CPU models with -IBRS suffix.
  336. Must be explicitly turned on for Intel CPU models without -IBRS suffix.
  337. Requires an updated host CPU microcode (intel-microcode >= 20180425).
  338. +
  339. * 'ssbd'
  340. +
  341. Required to enable the Spectre V4 (CVE-2018-3639) fix. Not included by default in any Intel CPU model.
  342. Must be explicitly turned on for all Intel CPU models.
  343. Requires an updated host CPU microcode(intel-microcode >= 20180703).
  344. AMD processors
  345. ^^^^^^^^^^^^^^
  346. * 'ibpb'
  347. +
  348. Required to enable the Spectre v1 (CVE-2017-5753) and Spectre v2 (CVE-2017-5715) fix,
  349. in cases where retpolines are not sufficient.
  350. Included by default in AMD CPU models with -IBPB suffix.
  351. Must be explicitly turned on for AMD CPU models without -IBPB suffix.
  352. Requires the host CPU microcode to support this feature before it can be used for guest CPUs.
  353. * 'virt-ssbd'
  354. +
  355. Required to enable the Spectre v4 (CVE-2018-3639) fix.
  356. Not included by default in any AMD CPU model.
  357. Must be explicitly turned on for all AMD CPU models.
  358. This should be provided to guests, even if amd-ssbd is also provided, for maximum guest compatibility.
  359. Note that this must be explicitly enabled when when using the "host" cpu model,
  360. because this is a virtual feature which does not exist in the physical CPUs.
  361. * 'amd-ssbd'
  362. +
  363. Required to enable the Spectre v4 (CVE-2018-3639) fix.
  364. Not included by default in any AMD CPU model. Must be explicitly turned on for all AMD CPU models.
  365. This provides higher performance than virt-ssbd, therefore a host supporting this should always expose this to guests if possible.
  366. virt-ssbd should none the less also be exposed for maximum guest compatibility as some kernels only know about virt-ssbd.
  367. * 'amd-no-ssb'
  368. +
  369. Recommended to indicate the host is not vulnerable to Spectre V4 (CVE-2018-3639).
  370. Not included by default in any AMD CPU model.
  371. Future hardware generations of CPU will not be vulnerable to CVE-2018-3639,
  372. and thus the guest should be told not to enable its mitigations, by exposing amd-no-ssb.
  373. This is mutually exclusive with virt-ssbd and amd-ssbd.
  374. NUMA
  375. ^^^^
  376. You can also optionally emulate a *NUMA*
  377. footnote:[https://en.wikipedia.org/wiki/Non-uniform_memory_access] architecture
  378. in your VMs. The basics of the NUMA architecture mean that instead of having a
  379. global memory pool available to all your cores, the memory is spread into local
  380. banks close to each socket.
  381. This can bring speed improvements as the memory bus is not a bottleneck
  382. anymore. If your system has a NUMA architecture footnote:[if the command
  383. `numactl --hardware | grep available` returns more than one node, then your host
  384. system has a NUMA architecture] we recommend to activate the option, as this
  385. will allow proper distribution of the VM resources on the host system.
  386. This option is also required to hot-plug cores or RAM in a VM.
  387. If the NUMA option is used, it is recommended to set the number of sockets to
  388. the number of nodes of the host system.
  389. vCPU hot-plug
  390. ^^^^^^^^^^^^^
  391. Modern operating systems introduced the capability to hot-plug and, to a
  392. certain extent, hot-unplug CPUs in a running systems. Virtualisation allows us
  393. to avoid a lot of the (physical) problems real hardware can cause in such
  394. scenarios.
  395. Still, this is a rather new and complicated feature, so its use should be
  396. restricted to cases where its absolutely needed. Most of the functionality can
  397. be replicated with other, well tested and less complicated, features, see
  398. xref:qm_cpu_resource_limits[Resource Limits].
  399. In {pve} the maximal number of plugged CPUs is always `cores * sockets`.
  400. To start a VM with less than this total core count of CPUs you may use the
  401. *vpus* setting, it denotes how many vCPUs should be plugged in at VM start.
  402. Currently only this feature is only supported on Linux, a kernel newer than 3.10
  403. is needed, a kernel newer than 4.7 is recommended.
  404. You can use a udev rule as follow to automatically set new CPUs as online in
  405. the guest:
  406. ----
  407. SUBSYSTEM=="cpu", ACTION=="add", TEST=="online", ATTR{online}=="0", ATTR{online}="1"
  408. ----
  409. Save this under /etc/udev/rules.d/ as a file ending in `.rules`.
  410. Note: CPU hot-remove is machine dependent and requires guest cooperation.
  411. The deletion command does not guarantee CPU removal to actually happen,
  412. typically it's a request forwarded to guest using target dependent mechanism,
  413. e.g., ACPI on x86/amd64.
  414. [[qm_memory]]
  415. Memory
  416. ~~~~~~
  417. For each VM you have the option to set a fixed size memory or asking
  418. {pve} to dynamically allocate memory based on the current RAM usage of the
  419. host.
  420. .Fixed Memory Allocation
  421. [thumbnail="screenshot/gui-create-vm-memory.png"]
  422. When setting memory and minimum memory to the same amount
  423. {pve} will simply allocate what you specify to your VM.
  424. Even when using a fixed memory size, the ballooning device gets added to the
  425. VM, because it delivers useful information such as how much memory the guest
  426. really uses.
  427. In general, you should leave *ballooning* enabled, but if you want to disable
  428. it (e.g. for debugging purposes), simply uncheck
  429. *Ballooning Device* or set
  430. balloon: 0
  431. in the configuration.
  432. .Automatic Memory Allocation
  433. // see autoballoon() in pvestatd.pm
  434. When setting the minimum memory lower than memory, {pve} will make sure that the
  435. minimum amount you specified is always available to the VM, and if RAM usage on
  436. the host is below 80%, will dynamically add memory to the guest up to the
  437. maximum memory specified.
  438. When the host is running low on RAM, the VM will then release some memory
  439. back to the host, swapping running processes if needed and starting the oom
  440. killer in last resort. The passing around of memory between host and guest is
  441. done via a special `balloon` kernel driver running inside the guest, which will
  442. grab or release memory pages from the host.
  443. footnote:[A good explanation of the inner workings of the balloon driver can be found here https://rwmj.wordpress.com/2010/07/17/virtio-balloon/]
  444. When multiple VMs use the autoallocate facility, it is possible to set a
  445. *Shares* coefficient which indicates the relative amount of the free host memory
  446. that each VM should take. Suppose for instance you have four VMs, three of them
  447. running an HTTP server and the last one is a database server. To cache more
  448. database blocks in the database server RAM, you would like to prioritize the
  449. database VM when spare RAM is available. For this you assign a Shares property
  450. of 3000 to the database VM, leaving the other VMs to the Shares default setting
  451. of 1000. The host server has 32GB of RAM, and is currently using 16GB, leaving 32
  452. * 80/100 - 16 = 9GB RAM to be allocated to the VMs. The database VM will get 9 *
  453. 3000 / (3000 + 1000 + 1000 + 1000) = 4.5 GB extra RAM and each HTTP server will
  454. get 1.5 GB.
  455. All Linux distributions released after 2010 have the balloon kernel driver
  456. included. For Windows OSes, the balloon driver needs to be added manually and can
  457. incur a slowdown of the guest, so we don't recommend using it on critical
  458. systems.
  459. // see https://forum.proxmox.com/threads/solved-hyper-threading-vs-no-hyper-threading-fixed-vs-variable-memory.20265/
  460. When allocating RAM to your VMs, a good rule of thumb is always to leave 1GB
  461. of RAM available to the host.
  462. [[qm_network_device]]
  463. Network Device
  464. ~~~~~~~~~~~~~~
  465. [thumbnail="screenshot/gui-create-vm-network.png"]
  466. Each VM can have many _Network interface controllers_ (NIC), of four different
  467. types:
  468. * *Intel E1000* is the default, and emulates an Intel Gigabit network card.
  469. * the *VirtIO* paravirtualized NIC should be used if you aim for maximum
  470. performance. Like all VirtIO devices, the guest OS should have the proper driver
  471. installed.
  472. * the *Realtek 8139* emulates an older 100 MB/s network card, and should
  473. only be used when emulating older operating systems ( released before 2002 )
  474. * the *vmxnet3* is another paravirtualized device, which should only be used
  475. when importing a VM from another hypervisor.
  476. {pve} will generate for each NIC a random *MAC address*, so that your VM is
  477. addressable on Ethernet networks.
  478. The NIC you added to the VM can follow one of two different models:
  479. * in the default *Bridged mode* each virtual NIC is backed on the host by a
  480. _tap device_, ( a software loopback device simulating an Ethernet NIC ). This
  481. tap device is added to a bridge, by default vmbr0 in {pve}. In this mode, VMs
  482. have direct access to the Ethernet LAN on which the host is located.
  483. * in the alternative *NAT mode*, each virtual NIC will only communicate with
  484. the Qemu user networking stack, where a built-in router and DHCP server can
  485. provide network access. This built-in DHCP will serve addresses in the private
  486. 10.0.2.0/24 range. The NAT mode is much slower than the bridged mode, and
  487. should only be used for testing. This mode is only available via CLI or the API,
  488. but not via the WebUI.
  489. You can also skip adding a network device when creating a VM by selecting *No
  490. network device*.
  491. .Multiqueue
  492. If you are using the VirtIO driver, you can optionally activate the
  493. *Multiqueue* option. This option allows the guest OS to process networking
  494. packets using multiple virtual CPUs, providing an increase in the total number
  495. of packets transferred.
  496. //http://blog.vmsplice.net/2011/09/qemu-internals-vhost-architecture.html
  497. When using the VirtIO driver with {pve}, each NIC network queue is passed to the
  498. host kernel, where the queue will be processed by a kernel thread spawned by the
  499. vhost driver. With this option activated, it is possible to pass _multiple_
  500. network queues to the host kernel for each NIC.
  501. //https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/Virtualization_Tuning_and_Optimization_Guide/sect-Virtualization_Tuning_Optimization_Guide-Networking-Techniques.html#sect-Virtualization_Tuning_Optimization_Guide-Networking-Multi-queue_virtio-net
  502. When using Multiqueue, it is recommended to set it to a value equal
  503. to the number of Total Cores of your guest. You also need to set in
  504. the VM the number of multi-purpose channels on each VirtIO NIC with the ethtool
  505. command:
  506. `ethtool -L ens1 combined X`
  507. where X is the number of the number of vcpus of the VM.
  508. You should note that setting the Multiqueue parameter to a value greater
  509. than one will increase the CPU load on the host and guest systems as the
  510. traffic increases. We recommend to set this option only when the VM has to
  511. process a great number of incoming connections, such as when the VM is running
  512. as a router, reverse proxy or a busy HTTP server doing long polling.
  513. [[qm_display]]
  514. Display
  515. ~~~~~~~
  516. QEMU can virtualize a few types of VGA hardware. Some examples are:
  517. * *std*, the default, emulates a card with Bochs VBE extensions.
  518. * *cirrus*, this was once the default, it emulates a very old hardware module
  519. with all its problems. This display type should only be used if really
  520. necessary footnote:[https://www.kraxel.org/blog/2014/10/qemu-using-cirrus-considered-harmful/
  521. qemu: using cirrus considered harmful], e.g., if using Windows XP or earlier
  522. * *vmware*, is a VMWare SVGA-II compatible adapter.
  523. * *qxl*, is the QXL paravirtualized graphics card. Selecting this also
  524. enables https://www.spice-space.org/[SPICE] (a remote viewer protocol) for the
  525. VM.
  526. You can edit the amount of memory given to the virtual GPU, by setting
  527. the 'memory' option. This can enable higher resolutions inside the VM,
  528. especially with SPICE/QXL.
  529. As the memory is reserved by display device, selecting Multi-Monitor mode
  530. for SPICE (e.g., `qxl2` for dual monitors) has some implications:
  531. * Windows needs a device for each monitor, so if your 'ostype' is some
  532. version of Windows, {pve} gives the VM an extra device per monitor.
  533. Each device gets the specified amount of memory.
  534. * Linux VMs, can always enable more virtual monitors, but selecting
  535. a Multi-Monitor mode multiplies the memory given to the device with
  536. the number of monitors.
  537. Selecting `serialX` as display 'type' disables the VGA output, and redirects
  538. the Web Console to the selected serial port. A configured display 'memory'
  539. setting will be ignored in that case.
  540. [[qm_usb_passthrough]]
  541. USB Passthrough
  542. ~~~~~~~~~~~~~~~
  543. There are two different types of USB passthrough devices:
  544. * Host USB passthrough
  545. * SPICE USB passthrough
  546. Host USB passthrough works by giving a VM a USB device of the host.
  547. This can either be done via the vendor- and product-id, or
  548. via the host bus and port.
  549. The vendor/product-id looks like this: *0123:abcd*,
  550. where *0123* is the id of the vendor, and *abcd* is the id
  551. of the product, meaning two pieces of the same usb device
  552. have the same id.
  553. The bus/port looks like this: *1-2.3.4*, where *1* is the bus
  554. and *2.3.4* is the port path. This represents the physical
  555. ports of your host (depending of the internal order of the
  556. usb controllers).
  557. If a device is present in a VM configuration when the VM starts up,
  558. but the device is not present in the host, the VM can boot without problems.
  559. As soon as the device/port is available in the host, it gets passed through.
  560. WARNING: Using this kind of USB passthrough means that you cannot move
  561. a VM online to another host, since the hardware is only available
  562. on the host the VM is currently residing.
  563. The second type of passthrough is SPICE USB passthrough. This is useful
  564. if you use a SPICE client which supports it. If you add a SPICE USB port
  565. to your VM, you can passthrough a USB device from where your SPICE client is,
  566. directly to the VM (for example an input device or hardware dongle).
  567. [[qm_bios_and_uefi]]
  568. BIOS and UEFI
  569. ~~~~~~~~~~~~~
  570. In order to properly emulate a computer, QEMU needs to use a firmware.
  571. Which, on common PCs often known as BIOS or (U)EFI, is executed as one of the
  572. first steps when booting a VM. It is responsible for doing basic hardware
  573. initialization and for providing an interface to the firmware and hardware for
  574. the operating system. By default QEMU uses *SeaBIOS* for this, which is an
  575. open-source, x86 BIOS implementation. SeaBIOS is a good choice for most
  576. standard setups.
  577. There are, however, some scenarios in which a BIOS is not a good firmware
  578. to boot from, e.g. if you want to do VGA passthrough. footnote:[Alex Williamson has a very good blog entry about this.
  579. http://vfio.blogspot.co.at/2014/08/primary-graphics-assignment-without-vga.html]
  580. In such cases, you should rather use *OVMF*, which is an open-source UEFI implementation. footnote:[See the OVMF Project http://www.tianocore.org/ovmf/]
  581. If you want to use OVMF, there are several things to consider:
  582. In order to save things like the *boot order*, there needs to be an EFI Disk.
  583. This disk will be included in backups and snapshots, and there can only be one.
  584. You can create such a disk with the following command:
  585. qm set <vmid> -efidisk0 <storage>:1,format=<format>
  586. Where *<storage>* is the storage where you want to have the disk, and
  587. *<format>* is a format which the storage supports. Alternatively, you can
  588. create such a disk through the web interface with 'Add' -> 'EFI Disk' in the
  589. hardware section of a VM.
  590. When using OVMF with a virtual display (without VGA passthrough),
  591. you need to set the client resolution in the OVMF menu(which you can reach
  592. with a press of the ESC button during boot), or you have to choose
  593. SPICE as the display type.
  594. [[qm_ivshmem]]
  595. Inter-VM shared memory
  596. ~~~~~~~~~~~~~~~~~~~~~~
  597. You can add an Inter-VM shared memory device (`ivshmem`), which allows one to
  598. share memory between the host and a guest, or also between multiple guests.
  599. To add such a device, you can use `qm`:
  600. qm set <vmid> -ivshmem size=32,name=foo
  601. Where the size is in MiB. The file will be located under
  602. `/dev/shm/pve-shm-$name` (the default name is the vmid).
  603. NOTE: Currently the device will get deleted as soon as any VM using it got
  604. shutdown or stopped. Open connections will still persist, but new connections
  605. to the exact same device cannot be made anymore.
  606. A use case for such a device is the Looking Glass
  607. footnote:[Looking Glass: https://looking-glass.hostfission.com/] project,
  608. which enables high performance, low-latency display mirroring between
  609. host and guest.
  610. [[qm_audio_device]]
  611. Audio Device
  612. ~~~~~~~~~~~~
  613. To add an audio device run the following command:
  614. ----
  615. qm set <vmid> -audio0 device=<device>
  616. ----
  617. Supported audio devices are:
  618. * `ich9-intel-hda`: Intel HD Audio Controller, emulates ICH9
  619. * `intel-hda`: Intel HD Audio Controller, emulates ICH6
  620. * `AC97`: Audio Codec '97, useful for older operating systems like Windows XP
  621. NOTE: The audio device works only in combination with SPICE. Remote protocols
  622. like Microsoft's RDP have options to play sound. To use the physical audio
  623. device of the host use device passthrough (see
  624. xref:qm_pci_passthrough[PCI Passthrough] and
  625. xref:qm_usb_passthrough[USB Passthrough]).
  626. [[qm_startup_and_shutdown]]
  627. Automatic Start and Shutdown of Virtual Machines
  628. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  629. After creating your VMs, you probably want them to start automatically
  630. when the host system boots. For this you need to select the option 'Start at
  631. boot' from the 'Options' Tab of your VM in the web interface, or set it with
  632. the following command:
  633. qm set <vmid> -onboot 1
  634. .Start and Shutdown Order
  635. [thumbnail="screenshot/gui-qemu-edit-start-order.png"]
  636. In some case you want to be able to fine tune the boot order of your
  637. VMs, for instance if one of your VM is providing firewalling or DHCP
  638. to other guest systems. For this you can use the following
  639. parameters:
  640. * *Start/Shutdown order*: Defines the start order priority. E.g. set it to 1 if
  641. you want the VM to be the first to be started. (We use the reverse startup
  642. order for shutdown, so a machine with a start order of 1 would be the last to
  643. be shut down). If multiple VMs have the same order defined on a host, they will
  644. additionally be ordered by 'VMID' in ascending order.
  645. * *Startup delay*: Defines the interval between this VM start and subsequent
  646. VMs starts . E.g. set it to 240 if you want to wait 240 seconds before starting
  647. other VMs.
  648. * *Shutdown timeout*: Defines the duration in seconds {pve} should wait
  649. for the VM to be offline after issuing a shutdown command.
  650. By default this value is set to 180, which means that {pve} will issue a
  651. shutdown request and wait 180 seconds for the machine to be offline. If
  652. the machine is still online after the timeout it will be stopped forcefully.
  653. NOTE: VMs managed by the HA stack do not follow the 'start on boot' and
  654. 'boot order' options currently. Those VMs will be skipped by the startup and
  655. shutdown algorithm as the HA manager itself ensures that VMs get started and
  656. stopped.
  657. Please note that machines without a Start/Shutdown order parameter will always
  658. start after those where the parameter is set. Further, this parameter can only
  659. be enforced between virtual machines running on the same host, not
  660. cluster-wide.
  661. [[qm_spice_enhancements]]
  662. SPICE Enhancements
  663. ~~~~~~~~~~~~~~~~~~
  664. SPICE Enhancements are optional features that can improve the remote viewer
  665. experience.
  666. To enable them via the GUI go to the *Options* panel of the virtual machine. Run
  667. the following command to enable them via the CLI:
  668. ----
  669. qm set <vmid> -spice_enhancements foldersharing=1,videostreaming=all
  670. ----
  671. NOTE: To use these features the <<qm_display,*Display*>> of the virtual machine
  672. must be set to SPICE (qxl).
  673. Folder Sharing
  674. ^^^^^^^^^^^^^^
  675. Share a local folder with the guest. The `spice-webdavd` daemon needs to be
  676. installed in the guest. It makes the shared folder available through a local
  677. WebDAV server located at http://localhost:9843.
  678. For Windows guests the installer for the 'Spice WebDAV daemon' can be downloaded
  679. from the
  680. https://www.spice-space.org/download.html#windows-binaries[official SPICE website].
  681. Most Linux distributions have a package called `spice-webdavd` that can be
  682. installed.
  683. To share a folder in Virt-Viewer (Remote Viewer) go to 'File -> Preferences'.
  684. Select the folder to share and then enable the checkbox.
  685. NOTE: Folder sharing currently only works in the Linux version of Virt-Viewer.
  686. CAUTION: Experimental! Currently this feature does not work reliably.
  687. Video Streaming
  688. ^^^^^^^^^^^^^^^
  689. Fast refreshing areas are encoded into a video stream. Two options exist:
  690. * *all*: Any fast refreshing area will be encoded into a video stream.
  691. * *filter*: Additional filters are used to decide if video streaming should be
  692. used (currently only small window surfaces are skipped).
  693. A general recommendation if video streaming should be enabled and which option
  694. to choose from cannot be given. Your mileage may vary depending on the specific
  695. circumstances.
  696. Troubleshooting
  697. ^^^^^^^^^^^^^^^
  698. .Shared folder does not show up
  699. Make sure the WebDAV service is enabled and running in the guest. On Windows it
  700. is called 'Spice webdav proxy'. In Linux the name is 'spice-webdavd' but can be
  701. different depending on the distribution.
  702. If the service is running, check the WebDAV server by opening
  703. http://localhost:9843 in a browser in the guest.
  704. It can help to restart the SPICE session.
  705. [[qm_migration]]
  706. Migration
  707. ---------
  708. [thumbnail="screenshot/gui-qemu-migrate.png"]
  709. If you have a cluster, you can migrate your VM to another host with
  710. qm migrate <vmid> <target>
  711. There are generally two mechanisms for this
  712. * Online Migration (aka Live Migration)
  713. * Offline Migration
  714. Online Migration
  715. ~~~~~~~~~~~~~~~~
  716. When your VM is running and it has no local resources defined (such as disks
  717. on local storage, passed through devices, etc.) you can initiate a live
  718. migration with the -online flag.
  719. How it works
  720. ^^^^^^^^^^^^
  721. This starts a Qemu Process on the target host with the 'incoming' flag, which
  722. means that the process starts and waits for the memory data and device states
  723. from the source Virtual Machine (since all other resources, e.g. disks,
  724. are shared, the memory content and device state are the only things left
  725. to transmit).
  726. Once this connection is established, the source begins to send the memory
  727. content asynchronously to the target. If the memory on the source changes,
  728. those sections are marked dirty and there will be another pass of sending data.
  729. This happens until the amount of data to send is so small that it can
  730. pause the VM on the source, send the remaining data to the target and start
  731. the VM on the target in under a second.
  732. Requirements
  733. ^^^^^^^^^^^^
  734. For Live Migration to work, there are some things required:
  735. * The VM has no local resources (e.g. passed through devices, local disks, etc.)
  736. * The hosts are in the same {pve} cluster.
  737. * The hosts have a working (and reliable) network connection.
  738. * The target host must have the same or higher versions of the
  739. {pve} packages. (It *might* work the other way, but this is never guaranteed)
  740. Offline Migration
  741. ~~~~~~~~~~~~~~~~~
  742. If you have local resources, you can still offline migrate your VMs,
  743. as long as all disk are on storages, which are defined on both hosts.
  744. Then the migration will copy the disk over the network to the target host.
  745. [[qm_copy_and_clone]]
  746. Copies and Clones
  747. -----------------
  748. [thumbnail="screenshot/gui-qemu-full-clone.png"]
  749. VM installation is usually done using an installation media (CD-ROM)
  750. from the operation system vendor. Depending on the OS, this can be a
  751. time consuming task one might want to avoid.
  752. An easy way to deploy many VMs of the same type is to copy an existing
  753. VM. We use the term 'clone' for such copies, and distinguish between
  754. 'linked' and 'full' clones.
  755. Full Clone::
  756. The result of such copy is an independent VM. The
  757. new VM does not share any storage resources with the original.
  758. +
  759. It is possible to select a *Target Storage*, so one can use this to
  760. migrate a VM to a totally different storage. You can also change the
  761. disk image *Format* if the storage driver supports several formats.
  762. +
  763. NOTE: A full clone needs to read and copy all VM image data. This is
  764. usually much slower than creating a linked clone.
  765. +
  766. Some storage types allows to copy a specific *Snapshot*, which
  767. defaults to the 'current' VM data. This also means that the final copy
  768. never includes any additional snapshots from the original VM.
  769. Linked Clone::
  770. Modern storage drivers support a way to generate fast linked
  771. clones. Such a clone is a writable copy whose initial contents are the
  772. same as the original data. Creating a linked clone is nearly
  773. instantaneous, and initially consumes no additional space.
  774. +
  775. They are called 'linked' because the new image still refers to the
  776. original. Unmodified data blocks are read from the original image, but
  777. modification are written (and afterwards read) from a new
  778. location. This technique is called 'Copy-on-write'.
  779. +
  780. This requires that the original volume is read-only. With {pve} one
  781. can convert any VM into a read-only <<qm_templates, Template>>). Such
  782. templates can later be used to create linked clones efficiently.
  783. +
  784. NOTE: You cannot delete an original template while linked clones
  785. exist.
  786. +
  787. It is not possible to change the *Target storage* for linked clones,
  788. because this is a storage internal feature.
  789. The *Target node* option allows you to create the new VM on a
  790. different node. The only restriction is that the VM is on shared
  791. storage, and that storage is also available on the target node.
  792. To avoid resource conflicts, all network interface MAC addresses get
  793. randomized, and we generate a new 'UUID' for the VM BIOS (smbios1)
  794. setting.
  795. [[qm_templates]]
  796. Virtual Machine Templates
  797. -------------------------
  798. One can convert a VM into a Template. Such templates are read-only,
  799. and you can use them to create linked clones.
  800. NOTE: It is not possible to start templates, because this would modify
  801. the disk images. If you want to change the template, create a linked
  802. clone and modify that.
  803. VM Generation ID
  804. ----------------
  805. {pve} supports Virtual Machine Generation ID ('vmgenid') footnote:[Official
  806. 'vmgenid' Specification
  807. https://docs.microsoft.com/en-us/windows/desktop/hyperv_v2/virtual-machine-generation-identifier]
  808. for virtual machines.
  809. This can be used by the guest operating system to detect any event resulting
  810. in a time shift event, for example, restoring a backup or a snapshot rollback.
  811. When creating new VMs, a 'vmgenid' will be automatically generated and saved
  812. in its configuration file.
  813. To create and add a 'vmgenid' to an already existing VM one can pass the
  814. special value `1' to let {pve} autogenerate one or manually set the 'UUID'
  815. footnote:[Online GUID generator http://guid.one/] by using it as value,
  816. e.g.:
  817. ----
  818. qm set VMID -vmgenid 1
  819. qm set VMID -vmgenid 00000000-0000-0000-0000-000000000000
  820. ----
  821. NOTE: The initial addition of a 'vmgenid' device to an existing VM, may result
  822. in the same effects as a change on snapshot rollback, backup restore, etc., has
  823. as the VM can interpret this as generation change.
  824. In the rare case the 'vmgenid' mechanism is not wanted one can pass `0' for
  825. its value on VM creation, or retroactively delete the property in the
  826. configuration with:
  827. ----
  828. qm set VMID -delete vmgenid
  829. ----
  830. The most prominent use case for 'vmgenid' are newer Microsoft Windows
  831. operating systems, which use it to avoid problems in time sensitive or
  832. replicate services (e.g., databases, domain controller
  833. footnote:[https://docs.microsoft.com/en-us/windows-server/identity/ad-ds/get-started/virtual-dc/virtualized-domain-controller-architecture])
  834. on snapshot rollback, backup restore or a whole VM clone operation.
  835. Importing Virtual Machines and disk images
  836. ------------------------------------------
  837. A VM export from a foreign hypervisor takes usually the form of one or more disk
  838. images, with a configuration file describing the settings of the VM (RAM,
  839. number of cores). +
  840. The disk images can be in the vmdk format, if the disks come from
  841. VMware or VirtualBox, or qcow2 if the disks come from a KVM hypervisor.
  842. The most popular configuration format for VM exports is the OVF standard, but in
  843. practice interoperation is limited because many settings are not implemented in
  844. the standard itself, and hypervisors export the supplementary information
  845. in non-standard extensions.
  846. Besides the problem of format, importing disk images from other hypervisors
  847. may fail if the emulated hardware changes too much from one hypervisor to
  848. another. Windows VMs are particularly concerned by this, as the OS is very
  849. picky about any changes of hardware. This problem may be solved by
  850. installing the MergeIDE.zip utility available from the Internet before exporting
  851. and choosing a hard disk type of *IDE* before booting the imported Windows VM.
  852. Finally there is the question of paravirtualized drivers, which improve the
  853. speed of the emulated system and are specific to the hypervisor.
  854. GNU/Linux and other free Unix OSes have all the necessary drivers installed by
  855. default and you can switch to the paravirtualized drivers right after importing
  856. the VM. For Windows VMs, you need to install the Windows paravirtualized
  857. drivers by yourself.
  858. GNU/Linux and other free Unix can usually be imported without hassle. Note
  859. that we cannot guarantee a successful import/export of Windows VMs in all
  860. cases due to the problems above.
  861. Step-by-step example of a Windows OVF import
  862. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  863. Microsoft provides
  864. https://developer.microsoft.com/en-us/windows/downloads/virtual-machines/[Virtual Machines downloads]
  865. to get started with Windows development.We are going to use one of these
  866. to demonstrate the OVF import feature.
  867. Download the Virtual Machine zip
  868. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  869. After getting informed about the user agreement, choose the _Windows 10
  870. Enterprise (Evaluation - Build)_ for the VMware platform, and download the zip.
  871. Extract the disk image from the zip
  872. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  873. Using the `unzip` utility or any archiver of your choice, unpack the zip,
  874. and copy via ssh/scp the ovf and vmdk files to your {pve} host.
  875. Import the Virtual Machine
  876. ^^^^^^^^^^^^^^^^^^^^^^^^^^
  877. This will create a new virtual machine, using cores, memory and
  878. VM name as read from the OVF manifest, and import the disks to the +local-lvm+
  879. storage. You have to configure the network manually.
  880. qm importovf 999 WinDev1709Eval.ovf local-lvm
  881. The VM is ready to be started.
  882. Adding an external disk image to a Virtual Machine
  883. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  884. You can also add an existing disk image to a VM, either coming from a
  885. foreign hypervisor, or one that you created yourself.
  886. Suppose you created a Debian/Ubuntu disk image with the 'vmdebootstrap' tool:
  887. vmdebootstrap --verbose \
  888. --size 10GiB --serial-console \
  889. --grub --no-extlinux \
  890. --package openssh-server \
  891. --package avahi-daemon \
  892. --package qemu-guest-agent \
  893. --hostname vm600 --enable-dhcp \
  894. --customize=./copy_pub_ssh.sh \
  895. --sparse --image vm600.raw
  896. You can now create a new target VM for this image.
  897. qm create 600 --net0 virtio,bridge=vmbr0 --name vm600 --serial0 socket \
  898. --bootdisk scsi0 --scsihw virtio-scsi-pci --ostype l26
  899. Add the disk image as +unused0+ to the VM, using the storage +pvedir+:
  900. qm importdisk 600 vm600.raw pvedir
  901. Finally attach the unused disk to the SCSI controller of the VM:
  902. qm set 600 --scsi0 pvedir:600/vm-600-disk-1.raw
  903. The VM is ready to be started.
  904. ifndef::wiki[]
  905. include::qm-cloud-init.adoc[]
  906. endif::wiki[]
  907. ifndef::wiki[]
  908. include::qm-pci-passthrough.adoc[]
  909. endif::wiki[]
  910. Hookscripts
  911. -----------
  912. You can add a hook script to VMs with the config property `hookscript`.
  913. qm set 100 -hookscript local:snippets/hookscript.pl
  914. It will be called during various phases of the guests lifetime.
  915. For an example and documentation see the example script under
  916. `/usr/share/pve-docs/examples/guest-example-hookscript.pl`.
  917. Managing Virtual Machines with `qm`
  918. ------------------------------------
  919. qm is the tool to manage Qemu/Kvm virtual machines on {pve}. You can
  920. create and destroy virtual machines, and control execution
  921. (start/stop/suspend/resume). Besides that, you can use qm to set
  922. parameters in the associated config file. It is also possible to
  923. create and delete virtual disks.
  924. CLI Usage Examples
  925. ~~~~~~~~~~~~~~~~~~
  926. Using an iso file uploaded on the 'local' storage, create a VM
  927. with a 4 GB IDE disk on the 'local-lvm' storage
  928. qm create 300 -ide0 local-lvm:4 -net0 e1000 -cdrom local:iso/proxmox-mailgateway_2.1.iso
  929. Start the new VM
  930. qm start 300
  931. Send a shutdown request, then wait until the VM is stopped.
  932. qm shutdown 300 && qm wait 300
  933. Same as above, but only wait for 40 seconds.
  934. qm shutdown 300 && qm wait 300 -timeout 40
  935. [[qm_configuration]]
  936. Configuration
  937. -------------
  938. VM configuration files are stored inside the Proxmox cluster file
  939. system, and can be accessed at `/etc/pve/qemu-server/<VMID>.conf`.
  940. Like other files stored inside `/etc/pve/`, they get automatically
  941. replicated to all other cluster nodes.
  942. NOTE: VMIDs < 100 are reserved for internal purposes, and VMIDs need to be
  943. unique cluster wide.
  944. .Example VM Configuration
  945. ----
  946. cores: 1
  947. sockets: 1
  948. memory: 512
  949. name: webmail
  950. ostype: l26
  951. bootdisk: virtio0
  952. net0: e1000=EE:D2:28:5F:B6:3E,bridge=vmbr0
  953. virtio0: local:vm-100-disk-1,size=32G
  954. ----
  955. Those configuration files are simple text files, and you can edit them
  956. using a normal text editor (`vi`, `nano`, ...). This is sometimes
  957. useful to do small corrections, but keep in mind that you need to
  958. restart the VM to apply such changes.
  959. For that reason, it is usually better to use the `qm` command to
  960. generate and modify those files, or do the whole thing using the GUI.
  961. Our toolkit is smart enough to instantaneously apply most changes to
  962. running VM. This feature is called "hot plug", and there is no
  963. need to restart the VM in that case.
  964. File Format
  965. ~~~~~~~~~~~
  966. VM configuration files use a simple colon separated key/value
  967. format. Each line has the following format:
  968. -----
  969. # this is a comment
  970. OPTION: value
  971. -----
  972. Blank lines in those files are ignored, and lines starting with a `#`
  973. character are treated as comments and are also ignored.
  974. [[qm_snapshots]]
  975. Snapshots
  976. ~~~~~~~~~
  977. When you create a snapshot, `qm` stores the configuration at snapshot
  978. time into a separate snapshot section within the same configuration
  979. file. For example, after creating a snapshot called ``testsnapshot'',
  980. your configuration file will look like this:
  981. .VM configuration with snapshot
  982. ----
  983. memory: 512
  984. swap: 512
  985. parent: testsnaphot
  986. ...
  987. [testsnaphot]
  988. memory: 512
  989. swap: 512
  990. snaptime: 1457170803
  991. ...
  992. ----
  993. There are a few snapshot related properties like `parent` and
  994. `snaptime`. The `parent` property is used to store the parent/child
  995. relationship between snapshots. `snaptime` is the snapshot creation
  996. time stamp (Unix epoch).
  997. [[qm_options]]
  998. Options
  999. ~~~~~~~
  1000. include::qm.conf.5-opts.adoc[]
  1001. Locks
  1002. -----
  1003. Online migrations, snapshots and backups (`vzdump`) set a lock to
  1004. prevent incompatible concurrent actions on the affected VMs. Sometimes
  1005. you need to remove such a lock manually (e.g., after a power failure).
  1006. qm unlock <vmid>
  1007. CAUTION: Only do that if you are sure the action which set the lock is
  1008. no longer running.
  1009. ifdef::wiki[]
  1010. See Also
  1011. ~~~~~~~~
  1012. * link:/wiki/Cloud-Init_Support[Cloud-Init Support]
  1013. endif::wiki[]
  1014. ifdef::manvolnum[]
  1015. Files
  1016. ------
  1017. `/etc/pve/qemu-server/<VMID>.conf`::
  1018. Configuration file for the VM '<VMID>'.
  1019. include::pve-copyright.adoc[]
  1020. endif::manvolnum[]